On The Approach to Nanoscale Robots: Understanding the Relationship between Nanomotor's Architecture and Active Motion

Nanomotors with active motion have recently attracted wide attention. Researchers have developed a series of nanomotors with different architectures and propulsion mechanisms and have also explored the application prospects of nanomotors in many fields. Yet, the present nanomotors with simple component and motion behaviors are still far from the ultimate goal of nanoscale robots with controlled sophisticated motions. With this goal in mind, researchers are no longer satisfied with just making nanoparticles move actively, but looking forward to achieving precise control of the motion behavior of nanomotors. With the advance in nanofabrication techniques, tuning the nanomotors’ motion by modulating the structure of nanoparticles has become a widely welcomed approach. This article reviews the self‐propelling mechanisms and the various synthesis methods of nanomotors. It is also systematically summarized how the composition, surface properties, size, and morphology of nanoparticles affect their motion behavior. Some outstanding works are highlighted, the shortcomings, challenges, and opportunities in the field are also discussed. It is believed that a thorough review of the architecture–motion relationships of nanomotors is beneficial for the development of the field toward nanoscale robots.


Understanding the Driving Force of the Nanomotors
As mentioned earlier, the key to having the active motion of nanoparticles is to be able to apply a continuous asymmetric force on the nanoparticles. In the currently prepared nanomotors, there are two main sources of this force: the mass and energy exchange between the nanoparticles and the surrounding solution (often referred to as self-propulsion), and the force applied directly to the nanoparticles externally. In this part, we carefully categorize these mechanisms, so as to better understand how nanoparticles achieve active motion and how to control the motion. In this section, we only focus on why nanoparticles move and how to fabricate and control nanomotors with active motion are reviewed in the following sections.

Brownian Movement
Nanoparticles in solution are constantly in irregular Brownian motion. Due to the thermal motion of solvent molecules, the particles are subjected to collisions with solvent molecules from all directions, and the motion of the particles constantly changes direction, resulting in irregular Brownian motion of the particles. Brownian motion is an intrinsic characteristic of microscopic particles, but different sizes of particles are affected by Brownian motion very differently. For "micromotors" with diameters of more than one micron, the Brownian motion is extremely weak, and it can be regarded as stationary when observed through an optical microscope. Therefore, for micromotors, the Brownian can be ignored, and the rate and direction of motion can be controlled very precisely by controlling the external environments. For motors with a diameter ranging from a few nanometers to hundreds of nanometers, i.e., "nanomotors", their true morphologies are difficult to observe by an optical microscope. Only the corresponding position of the particles can be known through the reflected light spot of the nanomotors using a dark field microscope. In this case, the nanomotors are found to be in irregular Brownian motions. For nanomotors, it is generally only possible to intensify or suppress the speed and radius of the motion, but not to make it linear.

Self-Propulsion
The driving force caused by the interaction of nanoparticles with the surrounding solution is also often called "self-propulsion" because it is generally caused by the physicochemical processes of the nanoparticles themselves. Nanoparticles' physicochemical processes with the surrounding result in localized asymmetric environments, which then induce anisotropic propulsion forces on the nanoparticles that make them move. The processes and mechanisms of such driving are complex, and researchers have proposed various mechanisms, including bubble driving, interfacial tension gradient driving, self-electrophoresis, Scheme 1. Schematic illustration of controlling the motion of nanomotors through nanoparticles' composition, surface properties, size, and morphologies.
www.advancedsciencenews.com www.advintellsyst.com self-diffusiophoresis, osmophoresis, thermophoresis, and so on. Based on the cause of the force on nanomotors, we have made a simplified classification of the self-propulsion mechanisms and divided them into four main types: the asymmetric interactions between the nanoparticles and the molecules in solution, the propulsion toward the nanoparticles by the solution flow, the asymmetric forces nanoparticles undertake from the localized electric field, and the bubble driving mechanism.

Asymmetric Molecular Interactions
Based on thermodynamic factors, molecules in solution (water molecules and solute molecules) are in constant Brownian motion, and they collide with the nanoparticle surfaces during their motion, creating interactions (attraction or repulsion). If the interactions from molecules on the nanoparticles are uniform in all directions, the nanoparticles are in passive Brownian motion with noncontrollable movements. However, if the interaction is stronger in a certain direction, that is, nanoparticle is subjected to a stronger force in one direction, this interaction force has an asymmetry that drives the nanoparticle into directional motion, which is often called diffusiophoresis ( Figure 1A). [22] This mechanism is often utilized to explain the chemotaxis of enzymes and molecules in gradient solutions. The interaction between enzymes and substrates is basically attraction forces because enzymes have a strong tendency of capturing the substrates for catalysis process, thus enzymes move toward the direction with higher substrate concentration ( Figure 1B). [23][24][25] Guha et al. demonstrated that Rh6G dyes chemotaxis toward high ficoll concentration, which can be attributed to the binding interactions between the dye and polymer. [26] Such anisotropic interaction-induced propulsion can also be able to be applied to nanoparticles. [27] For example, when enzymes are asymmetrically loaded on nanoparticles, they catalyze the substrate and thus generate a gradient of substrate and reaction product. Nanoparticles are thus propelled based on interactions with these two concentration gradients ( Figure 1A). [28,29] However, asymmetric interaction with solute molecules is not the most dominant force for propelling nanomotors in concentration gradient, as often other kinds of propulsions are accompanying this interaction and would often be stronger than pure molecular collision interactions, this will be discussed later. Temperature gradient also leads to asymmetric nanoparticlemolecule interactions. [30,31] When there is a temperature difference between the two sides of the nanoparticle, the water molecules on the higher temperature side move at a faster rate and give a stronger push to the nanoparticle when colliding with it, so the overall force on the nanoparticle is directed toward the low-temperature side. This phenomenon that nanoparticles move toward the low-temperature direction is often called thermophoresis, and when such temperature gradient is generated by the nanoparticle itself (photothermal effect for example), it is also named self-thermophoresis ( Figure 1C).

Localized Anisotropic Flow
For low Reynolds number swimmers, viscosity dominates the forces that make the particles move, [4] so the flow of liquid in Figure 1. Different propulsion mechanisms driving the active motion of self-propelled nanomotors. A) Diffusiophoresis of a nanoparticle due to concentration gradient induced asymmetric molecular interactions, molecules in the solution interact with nanoparticles, and the side with higher molecule concentration induces stronger interaction. B) Diffusiophoresis of an enzyme due to concentration gradient induced asymmetric molecular interactions, enzymes have a strong binding force toward certain molecules, thus move toward the direction with high molecule concentration. C) Thermophoresis of a nanoparticle due to temperature gradient induced asymmetric molecular interactions. Water molecules move faster in higher concentration region, thus colloid more with nanoparticles, generating a stronger force on the higher temperature side. D) Diffusiophoresis of a nanoparticle due to local asymmetric flow. Liquid flow around nanoparticles, generating forces on the nanoparticles and push them to move. E) Diffusiophoresis of a droplet due to the Marangoni mechanism. The different concentration of surfactants induces varied surface tension around nanoparticles, thus induce flow of liquid and thus propelling the nanoparticles. F) Diffusiophoresis of a semi-permeable nanoparticle based on concentration gradient induced asymmetric flow. Liquid flow both through and around the nanoparticles, thus the forces generated on the nanoparticles are complex and require specific investigation. the environment around the nanoparticles is critical to their motion. For systems such as blood flow, the nanoparticles move in the same direction as the liquid flow with passive motion and are motionless (apart from Brownian movement) with the blood flow as reference. In contrast, for a solution at rest, localized flowing of the solution is required for nanoparticles to move actively. In the current nanomotor research, this flow is generated by two main cases: concentration gradient and temperature gradient.
When a solute concentration gradient exists within a solution, osmotic pressure leads to the formation of a flow directing from the low concentration side to the high, propelling the nanoparticles ( Figure 1D). Such phenomenon has been observed using both electrolytes such as various salts [32] and nonelectrolytes (e.g., glucose), [33] and osmotic flow induced nanoparticle motion has also been widely studied. [34][35][36] For example, a Janus Ir&SiO 2 nanomotor with Ir modified on only one side of the SiO 2 nanosphere, Ir catalysis N 2 H 4 into N 2 , NH 3 and H 2 , and the products are concentrated on the side with Ir and not the other. These products create an osmotic pressure from the Ir-modified side toward the none-modified side, which leads to fluid flow across the nanomotor and thus propel the nanomotor. [37] Lots of researchers have also in turn utilized the passive movements of tracer nanoparticles to study the flow of water in complex environments. [38,39] Solvent flow due to temperature differences is also a common way to induce the motion of nanoparticles. In the presence of temperature differences, water flows from high to low temperatures, driving the particles. [40,41] As can be seen from the above descriptions, concentration/ temperature gradients cause anisotropic molecular interactions and asymmetric flow at the same time. Such case that multiple forces contribute to the propulsion of nanomotors is quite common, which requires very careful investigations. In the case of concentration gradient, the osmophoresis-induced propulsion of nanomotors is the main driving force, especially in the gradient of nonelectrolytes. [37,42,43] As for the case of temperature gradient, it is currently inconclusive whether molecule collision or convection dominates the thermophoresis of nanoparticles, with the majority of arguments in favor of the former.
Emulsion droplets consisting of amphiphilic surfactants and nonwater-soluble oils also exhibit self-propulsion properties in solutions where a concentration gradient of surfactant molecules is present. This is also a class of flow-induced movement, although due to the special dynamic surface and liquid exchange with surroundings, it is often singled out as the Marangoni mechanism ( Figure 1E). [44][45][46] Two sets of flows are generally presented at the same time around a droplet. One is the flow around the droplets due to the interfacial tension caused by the surfactant concentration difference. Emulsion droplets are in a dynamic state, and the surfactant coverage is inhomogeneous on the droplet surface, leading to tension along the interface. The other flow is the convection within the dynamic emulsion droplet, as it continues to fuse and precipitate micelles from surrounding solutions. These two flows together exert forces on the emulsion droplet to induce its self-propulsion toward the high-tension side. [47] Another issue that should be noted is that the direction of the propulsion force depends on the permeability of the nanoparticles and may be not the same as the fluid flow. This is especially crucial in concentration gradient-induced flow. For solid nanoparticles that are totally impermeable, nanoparticles are pushed along the direction of flow. [48] For semi-permeable nanoparticles, lipid nanoparticles that only solvent can penetrate inside, for example, the solvent flows both around and through the nanoparticles and lead to quite different propulsions. Many researchers have demonstrated that nanoparticles are pushed against the flow, regardless of the concentration inside the nanoparticles ( Figure 1F). [49,50] As for nanoparticles with large pores that both solvent and solute can flow within, the introduction of pores makes the system more complex, and systematic studies on how gradient concentration affects the propulsion force are yet to be done.

Localized Electric Field
The directional movement of colloidal particles in a dispersion medium toward the cathode or anode under the action of an applied electric field is called electrophoresis. Electrophoresis is widely used in many fields of analytical chemistry and biochemistry, and researchers have also applied the phenomenon of electrophoresis to the active motion of nanoparticles.
Besides externally applied, this electric field is mainly generated by two forms. One is when there is an electrolyte concentration gradient in the solution, the anions and cations flow from higher to lower concentrations. As a whole, because anions and cations coexist, the solution has a balanced charge. However, during the flow, anions and cations have different mass transfer rates due to the different hydration radii. They thus move at different speed, which in turn leads to the disruption of the local charge equilibrium and the generation of an electric potential field ( Figure 2A). [48,51,52] Once a localized electric potential field is generated, nanoparticles move toward one direction based on their surface charge. This is the main cause of nanoparticles' chemotaxis in solutions with the electrolyte concentration gradient.
When an electric field is generated in situ around a charged particle, this propelling mechanism is called self-electrophoresis. This usually originates from the redox reactions carried out on the surface of nanoparticles. When redox reactions occur at different parts of a particle surface, an ion concentration gradient is formed and hence a local electric field, leading to the motion of the object itself ( Figure 2B). Such redox reaction-induced selfelectrophoresis has been widely utilized as the propulsion force of nanomotors. The pioneer work of nanomotor from Prof. Sen & Mallouk group is based on this mechanism, where a Janus Au&Pt nanorod is self-propelled in an H 2 O 2 solution ( Figure 2C). [7,53] Oxidation of H 2 O 2 occurs at the Pt section, and reduction of H 2 O 2 and O 2 occurs at the Au section. As the oxidation process generates H þ (proton) and reduction consumes, a proton concentration gradient oriented from Pt to Au is formed. This asymmetric distribution results in the formation of a localized electric field. The negatively charged nanorod therefore moves toward the Pt side. Recently, Lyu et al. demonstrated that a H 2 O 2 fueled Janus Pt&SiO 2 nanomotor also moves through self-electrophoresis, rather than the often-believed O 2 gradient propelled mechanism ( Figure 2D). [54] The Janus Pt&SiO 2 nanomotors are fabricated by sputtering a Pt layer, and a thickness gradient is created to be thick at the pole and thin at the equator. The thicker region of the Pt film at a pole serves as the oxidation reaction site, while the reduction occurs in thinner region at the equator. This in turn generates a localized electric field and propels the nanomotors.

Bubble Propelling
Bubble propulsion is a special class of propulsion mechanism, as a bubble is created to interact with the nanomotors ( Figure 3A). Nanoparticles generate oxygen, [55] hydrogen, [56][57][58] nitric oxide, [59,60] and other gases in situ through catalytic reactions, forming bubbles. Metal-containing nanoparticles are often used for bubble propelling motors, for example, oxygen bubbles originate from the catalytic reaction of hydrogen peroxide by noble metals (Au & Pt) or that hydrogen bubbles can be formed through the replacement reaction of highly reactive metals (Zn and Mg) in acid or water. Gas bubbles can also be generated through enzyme-driven micromotors. As a typical example, Yang et al. loaded Catalase in mesopores containing UIO-66 MOF microparticles. In the presence of H 2 O 2 fuel, MOF motors exhibit jet-like propulsion enabled by enzymatic generation of oxygen bubbles. [61] Bubble propulsion has been widely studied, as the production of bubbles is very visual, they can be directly observed with an optical microscope.
The process of bubble propulsion can be divided into two processes. The first process is that a bubble is created clinging to the nanomotor and grows larger, pushing the nanomotor away from its original position. The second part is divided into two cases, one in which the bubble breaks away from the nanomotor and pushes it in the opposite direction. The nanomotor continuously generates bubbles in a linear motion ( Figure 3B). [55,58] In the other case, the bubble explodes and the instantaneous attribution of the solution generates a flow that pushes the nanomotor back to its original position (relative to before the bubble was generated), in this case, the nanomotor exhibits an oscillatory motion ( Figure 3C). [62][63][64] Both cases have been observed, and it is still not sure the factor that determines the behavior of the bubble.

External Propulsion
In the previous section, the main source of propelling force is the interaction between nanomotors and the solution. Because the nanoparticles interact directly with the solution, their forces and motions are also closely related to the nature of the solution. Therefore, for more complex environments such as living bodies and polluted water, the application of such nanomotors will be somewhat limited. There are also external forces that act directly on the nanoparticles to push them into moving. [65] Although this external force is not as simple as the nanoparticle "self-driven", it also has the advantage of stability. The main external forces are magnetic field, ultrasonic waves, and light.
Magnetic nanoparticles (Fe, Ni, Co, and corresponding metal oxides) can be propelled under a magnetic field. The direction of this motion can be modulated with the orientation of the magnetic field and is therefore widely used in the study of nanomotors with controllable trajectories. [66][67][68] However, the magnetic field is a gradient field, so although the direction of force is constant, the strength of the force varies with position, making it difficult to precisely control the rate of motion. In addition, the magnetic field strength decays quite significantly with distance, which makes the nanoparticles at a greater distance In the case of self-electrophoresis, redox-oxide reactions happen on different parts of the nanoparticles, the produced ions disrupt the local charge equilibrium, generates an electric potential field, and propel the nanoparticles. C) Scheme of a Janus Au&Pt nanoparticle's self-electrophoresis with H 2 O 2 as the fuel. Protons are generated on the Pt side, while protons are consumed on the Au side, thus a local electric field is formed. Reproduced with permission. [7] Copyright 2013 American Chemical Society. D) Scheme of a Janus Pt&SiO 2 nanoparticle's self-electrophoresis with H 2 O 2 as the fuel. Due to thickness variation of the Pt layer, the top thicker part acts as the cathode and consumes protons, while the thin part functions as anode and generates protons, the nanoparticles are thus propelled in this concentration gradient of proton. Reproduced with permission. [54] Copyright 2021 American Chemical Society.
www.advancedsciencenews.com www.advintellsyst.com underpowered. Many researchers have prepared magnetic nanomotors with a helical structure and provided a rotating magnetic field perpendicular to the radial direction ( Figure 4A). [69][70][71][72][73] At this point, the helical nanoparticles spin under the magnetic field, and the generated torque converts rotational motion to translational motion, which propels the nanomotors. [74,75] The rate of this motion can be precisely controlled by tunning the rotation frequency of the magnetic field, providing high controllability, so this strategy has been chosen for most of the recent magnetic nanomotors. For example, Prof. Mei's group prepared soft helix magnetic nanomotors, which under rotating magnetic field, may navigate through constricted microchannels, demonstrating new possibilities for biomedical operation at the micro and nanoscale ( Figure 4B). [76] Ultrasound/acoustic wave is often used in medicine for diagnosis and treatment, so ultrasound-driven nanomotors have a high prospect of application. The wave of ultrasound generates a force on the matters it interacts with, thus being able to propel nanomotors. Under ultrasonic waves, nanoparticles are propelled by the acoustic radiation forces and move toward pressure nodes, which have been utilized for nanomotors of all kinds and in a wild range of applications. [77] The first demonstration of controllable propelling of one single nanoparticle is made by Wang et al., which showed that metallic microrods can be controlled to propel, rotate, and assemble under ultrasonic waves ( Figure 4C,D). [78] They proposed that the microrods most likely are propelled by an asymmetric-induced acoustic pressure gradient, which is generally accepted as the propulsion mechanism of acoustic-based nanomotors.
Light may also generate force on nanoparticles. The optical tweezer is a Nobel Prize-winning technology that gives a force to micro and nanoparticles through the radiation pressure of  Reproduced with permission. [75] Copyright 2021 Elsevier and 2013 American Chemical Society. [77] www.advancedsciencenews.com www.advintellsyst.com light. [79] Maier et al. combined optical trapping and thermophoretic to achieve high accuracy injection of Au&Al 2 O 3 "nanopen" into cells. [80] They first used optical trapping to position the nanopen vertically on the surface of a cell, then utilized the thermophoretic propelling effect to achieve a high-accuracy injection of the nanopens into the cells. Although optical tweezer has been applied in many fields, it is mainly ap'plied to capture and control particles by high-power laser, which are not suitable for in vivo applications. Also, the force triggered directly by light, although having advantage in accuracy, is quite weak, therefore often used to "immobilize" rather than "propel" a nanoparticle. [81] At low irradiation intensity, nanomotors are more propelled by the photocatalytic and photothermal effects of visible-infrared lights. [9,11,82] In these cases, light irradiation is the "stimuli method", and thermophoresis and electrophoresis are the "propulsion mechanisms", and they are sometimes confused. Pacheco et al. fabricated a CdS&Fe 3 O 4 nanocrystals coincorporated polymer polycaprolactone micromotor. [83] When irradiated by visible light, charge transfer between CdS and Fe 3 O 4 leads to photocatalytic reactions with glucose-or peroxide-enriched media. The generation of a gradient of products around the micromotor endow the micromotor photopropelling ability. Voctor et al. synthesized WS 2 nanosheets, which under light irradiation, the light-induced heating of the WS 2 enables the nanosheets to function as micromotors with photophoretic motion. [84]

Design and Fabrication of Nanomotors
After understanding the mechanism of force motion of nanoparticles, we thus know that in order for nanomotors to perform "active motion", they need to be able to generate a local asymmetric environment (temperature gradient and concentration gradient) through physicochemical reactions. This imposes two requirements on nanoparticles: 1) nanomaterials should be able to perform certain physicochemical reactions. 2) Nanomaterials carry out physicochemical reactions with certain spatial asymmetry, which involves special synthesis methods such as selective modification, spraying, and anisotropic growth. Currently, the reactions used for generating propulsion include photothermal (Au), redox reactions (Pt, TiO 2 , etc.), enzymatic catalysis (catalase, glucose oxidase, lipase, urease, etc.), gas producing (Pt, Mg, Zn, etc.), and so on. Finding the active material is easy, while architecting them into asymmetric structure requires some fabrication techniques. In this section, we overview the strategy of constructing asymmetricity in nanoparticles to enable them as active nanomotors.

Intrinsic Anisotropy
Theoretically, asymmetric nanostructures must be prepared by special synthetic methods, but some are not, such as enzymatic modifications in the preparation of enzymatic nanomotors. During the preparation of nanomotors, enzymes with catalytic functions are modified on the surface of nanoparticles, which in turn catalyze the surrounding chemicals to form a concentration gradient or a local electric field. The enzymes should theoretically be distributed asymmetrically on the nanoparticle surface, thus to induce the asymmetric distribution of reaction products. [85] In practice, however, it has been shown that even if the enzymes are "uniformly" modified on the nanoparticle surface, the resulting enzymatic nanomotors still have a good ability to move autonomously. [86][87][88][89][90][91][92] Using stochastically optical reconstruction microscopy and molecular dynamics simulations, Prof. Sanchez and coworkers demonstrated that when enzymes are modified onto nanospheres, even when nanoparticles are uniformly covered with ligands for enzymes to attach, enzymes exhibit a patchy distributed fashion, which makes them intrinsically asymmetric. [93,94] They found that such a unevendistributing phenomenon exists for various kinds of nanospheres (PS & silica) and enzymes (urea and lipase). Different modification strategy affects the enzyme distribution, and most homogeneous distribution is achieved when enzymes are modified onto silica surface through hydrophobic interactions. But albeit with different strategies, the enzymes all exhibit patchy distributed fashion and thus the enzyme-modified nanospheres with autonomous propelling ability, revealing crucial factors for modulating enzyme-powered nanomotors. Such intrinsic anisotropy is also seen in other propulsion ways, as no nanoparticles are perfectly symmetric. [95] Although this intrinsic anisotropy helps simplify the fabrication process, it also means that the architecture cannot be precisely controlled, which hinders further research.

Template Deposition
Electrodeposition of metal into pores of commercial anodic alumina templates is a widely used method of fabricating rod-shaped nanomaterials. By controlling deposition time and sequences, that is, one metal is deposited first and other one followed to be deposited on top of it, multimetallic asymmetric nanorods can thus be obtained. For example, Prof. Sen and Mallouk group used this method to fabricate Janus Pt&Au nanorods ( Figure 5A). [7,96] A layer of Ag is first electroplated onto porous alumina membrane to block one side of the pore. Then Au and Pt are electrodeposited into the pores sequentially. After etching Ag with HNO 3 and the alumina membrane template with NaOH, Janus Au&Pt nanorods with well-defined asymmetric structure were obtained. Based on the selective catalytic reaction of H 2 O 2 in the Pt side and the generated localized electric field, the Janus Au&Pt nanorods can self-propel with H 2 O 2 as the fuel. This method enables control of component, size, and structure, thus is widely applied for fabricating nanomotors. [9,78,[97][98][99] But the diameter of the nanorods is a bit limited, as AAO templates usually have large pores (>100 nm), the Janus nanorods with micrometer length and diameter over 100 nm may not be suitable for biological applications. Another issue is that, because of the need to etch AAO template with NaOH, components stable in the alkaline environment is required, limiting the components that are optional.

Sputtering and Layer Deposition
Sputtering/layer deposition is also widely utilized methods to generate asymmetric nanomaterials. When nanospheres are dispersed on glass or silicon slides, their bottom side is blocked, creating a spatial asymmetry, thus enabling asymmetric surface modification through sputtering or layer deposition. A layer of  Figure 5B). [101] It should be noted that, in the case of Pt, active motion is not due to the asymmetric between Pt and MSN, but the thickness variation from the top to the equator due to sputtering. [54] Beside sputtering reactive components, this method also enables the deposition of inactive materials on active nanoparticles. SiO 2 can also be sputtered on NH 2 -modified MSNs to partially "block" the nanospheres. [102,103] In this way, enzymes are modified only on the unblocked side of the MSN, forming an asymmetric structure. When sputtering, matters are sputtered on the upper side of the nanospheres. One can also immerse the substrate with fixed nanoparticles into reaction medium, and all-but-the-bottom of the nanoparticles will be covered. For example, Wang et al. deposited a layer of polylactic acid-glycolic acid copolymer (PLGA) on Mg nanospheres, and PLGA covers the Mg nanospheres only leaving the bottom bare ( Figure 5C). When the asymmetric nanoparticles are dispersed in acid solution, Mg reacts with the acid to generate H 2 , allowing bubbling propulsion of the nanomotors. [56] Due to the simplicity of the method, this is currently the most used approach to fabricate Janus nanomotors. Yet still, problems remain, the biggest one is that it is rather difficult to control the ratio between deposited and nondeposited areas, making manipulation of morphology difficult. [104]

Emulsion Orientating
Nanoparticles can be immobilized on the surface of the oil-water interface, forming the Pickering emulsion. In this way, part of the nanoparticle is immersed in oil and the other part in water, and modifications can be done on either side to generate asymmetric structures. Kaang et al. utilized this strategy to fabricate Janus mSiO 2 @Fe 3 O 4 &PDA/Pt nanoparticles (PDA = polydopamine). mSiO 2 @Fe 3 O 4 nanospheres are locked at the paraffin surface, and PDA and Pt nanocrystals are sequentially modified from the water side. The Janus nanoparticles thus could be propelled by thermophoresis from PDA, electrophoresis from Pt, as well as by a magnetic field. [105] Several other works also used this strategy to fabricate Janus nanomotors. [106,107] Diez et al. used this method to fabricate Janus MSN&Pt nanoparticles, the active motion through electrophoresis enabled an enhanced drug delivery process. [108,109] On the contrary, nanoparticles immobilized on glass/silicon substrates that suffers from being fixed and the blockage percentage cannot be tuned, the problem with emulsion shielding strategy is that the blocking of nanoparticles by emulsion is a dynamic unstable form. Thus, the blocking percentage in the formed Janus structure is difficult to be controlled.
In addition to the emulsion shielding method, there is also an emulsion solvent evaporation method that can be used to prepare asymmetric nanoparticles. First, an oil-in-water emulsion is formed where functional nanocrystals and polymer are dispersed in the oil phase. During the evaporation of the oil phase, the nanocrystals are separated from the polymer, resulting in the formation of nanocrystal&polymer asymmetric microspheres. Jurado-Sanchez et al. described the synthesis of Janus graphene quantum dots (GOD) encapsulated polycaprolactone micromotor with magnetic Fe 3 O 4 NPs and Pt NPs asymmetrically encapsulated on one side. [110] This Janus micromomtor enable both bubble propulsion and magnetic actuation, thus functioning as a micromachines for the screening of complex urine and human serum samples. Similarly, Y. Li's group reported the emulsion solvent ecaporation method for a polycaprolactone&Pt/Fe 3 O 4 Janus micromotor. [111] The demonstrated that this micromotor enables autonomous movement in high ionic strength Figure 5. Schematic illustration and typical SEM image of A) the electrochemical synthesis of asymmetric metal rods, B) sputtering, and C) layered deposition methods for synthesis of Janus nanospheres. Reproduced with permission. [7] Copyright 2006 American Chemical Society, 2014 Wiley-VCH [98] and 2021 Wiley-VCH. [56] www.advancedsciencenews.com www.advintellsyst.com solution, remote control by magnetic, and sustained release of loaded drugs.

Anisotropic Nucleation
Normally, when certain precursor nucleates on one nanoparticle, core@shell structures are obtained through the homogeneous nucleation process. By tuning reaction conditions, the isotropic homogeneous nucleation process of nanomaterials may be altered into anisotropic nucleation, and asymmetric nanoparticles can be obtained. Li et al. demonstrated that raising the water/ethanol ratio during PMO (periodic mesoporous organosilica) nucleation on MSN surface, the interfacial energy between PMO and MSN is gradually increased. The increase in surface tension leads to anisotropic nucleation of PMO on MSNs, generating Janus nanoparticles MSN&PMO in which one MSN nanosphere is linked to one PMO nanocube. [112] This anisotropic nucleation of PMO can be applied to various kinds of nanomaterials, forming a series of NP&PMO Janus nanoparticles, thus widely used for the fabrication of Janus nanomaterials and nanomotors. [113][114][115] Because of the mesoporosity and easy functionalization of PMO, the obtained PMO-containing nanomotors have wide application potential. For example, Chen et al. fabricated Pt&PMO Janus nanoparticles through this anisotropic nucleation strategy ( Figure 6A). With H 2 O 2 as the fuel, the nanomotors exhibited active propelling behaviors which endowed the nanomotors enhanced tumor penetrating ability and excellent therapeutic performances. [116] Another kind of "nanobottle" shape asymmetric nanomotor is also widely used. [117] These nanobottles are fabricated through an emulsion templated method ( Figure 6B). [118][119][120] A water-in-oil emulsion is first prepared, and precursors (silica or polymer monomers) deposit on the surface of the droplets. During nanomaterial formation, water is extruded from the droplets, and the deposition of precursors follows the newly formed water-oil interfaces, eventually leading to a flask or bottle-like structure. Different from conventional nanospheres and nanorods, such nanobottles possess a single-opening cavity. When catalytic species are loaded within the cavity, the produced matters are released from the bottleneck, which produces chemical/thermal gradient very different from dense nanostructures. [29,55,121,122] Besides the above-mentioned two widely used strategies, there are various other methods of architecting asymmetric structures. Wilson and co-workers used controlled deformation of polymer vesicles to generate polymeric stomatocytes, these soft nanomotors with tunable morphologies enable stimuli-responsive control of the motion. [123][124][125][126] By exploiting the varied surface curvature between the edge and body of Au nanorods, Li et al. achieved selective growth of Pt at one end of Au nanorods with varying proportions of Pt shell coverage. They studied how coverage ratio affects the electrophoresis of bimetallic nanorods and found that the half-covered group reaches the highest diffusion coefficient. [127] Sanchez's group found that during the coating of mesoporous silica on PS nanospheres, the aggregated attachment between PS nanospheres during the sol-gel process leaves uncoated areas. Thus, after etching the PS, asymmetric hollow silica nanospheres with big-pore openings were obtained Figure 6. Examples of anisotropic nucleation induced asymmetric nanomotors. A) Schematic illustration of the anisotropic nucleation of PMO and the SEM image and propelling scheme of Janus Pt&PMO nanoparticles. Reproduced with permission. [111] Copyright 2021 American Chemical Society. B) Synthesis scheme, SEM, TEM images, and propelling scheme of silica nanobottles by extruding out a solvent from hollow particles. Reproduced with permission. [113] Copyright 2016 Wiley-VCH. C) Schematic illustration and the SEM image and propelling scheme of the hollow silica nanoparticles with an anisotropic big opening. Reproduced with permission. [123] Copyright 2019 Wiley-VCH.
www.advancedsciencenews.com www.advintellsyst.com ( Figure 6C). After modification of enzymes, such asymmetric hollow nanospheres turn into nanomotors with rapid active motion. [102,128] Fabricating asymmetric nanomotors through an anisotropic nucleation strategy can be a bit complex compared to the previously mentioned methods, but such a bottom-up approach is the key to realizing nanomotors with precisely tunned morphologies, which paves the way for further understanding of how architecture affects nanomotors' motion.

Biotemplating
A portion of organisms naturally have asymmetric morphology, and researchers can use them as templates to coat their surfaces with materials to form nanomotors with asymmetric morphology of the bureau. Some of the reported biological template nanomotors are red blood cells, [129] sperm, [130] pollen, [131] algae ( Figure 7A), [70] bacteria, [71] etc. Although the method is very simple, the resulting nanomotor morphology is limited by nature and cannot be more precisely tuned.

3D Printing
3D printing is present researchers' most powerful weapon toward nanoscale sophisticate architectures, and many nanomotors have been "printed". [132][133][134] These nanomotors have delicate morphologies beyond the touch of current bottom-up synthetic skills ( Figure 7B), thus enabling critical study on the self-propelling process. However, the resolution limit of the 3D printing technique impedes its fabrication of submicrometer materials, which hinders the usage of this technique on nanomotors for bio-applications. Some recent works report using novel strategies to extend the resolution of 3D printing into nanoscale, [135,136] we hope this might be used for the study of nanomotors in the near future.

Controlling the Motion of Nanomotors
In the course of studying nanomotors, researchers have found that some parameters significantly change the motion performance of nanomotors, including nanomotors' moving rate, direction, linearity, etc. These parameters include environmental factors [137][138][139] and structure factors. Considering that the environment surrounding the nanomotor is immutable in practical application scenarios (gastrointestinal, tumor microenvironment, etc.), the effect of nanoparticle structure on the motion behavior is thus of great interest. Researchers have explored the effects of different structures (compositions, surface properties, sizes, morphologies, etc.) on the motion of nanomotors. The mechanism of how architecture affects self-propelling has been investigated, and eventually, researchers have tried to achieve precise control of the motion performance of nanomotors by modulating the architecture. In particular, this field has developed rapidly in recent years with advances in synthetic Reproduced with permission. [69] Copyright 2021 American Chemical Society. B) SEM images of 3D printed microswimmers with various shapes. Reproduced with permission. [129] www.advancedsciencenews.com www.advintellsyst.com techniques and characterization tools. This section systematically reviews the progress in this area.

Effect of Composition on Nanomotors' Motion
The motion of nanomotors strongly depends on the active composition, and changes in composition have been found to have a significant impact on nanomotor motion. Lyu et al. found that for a Janus Pt&SiO 2 nanomotor, changing the configuration of Pt film into nanocrystals dramatically stops the motion of the nanomotors ( Figure 8A). [54] Through such structure-related different motion behavior, they demonstrated that Janus Pt&SiO 2 nanomotors in H 2 O 2 do not move through generating O 2 bubbles, but rather self-electrophoresis based on thickness gradient of Pt film ( Figure 2D). When an intact Pt film is changed into small Pt nanocrystals, the local electric field no longer exists, therefore the Janus Pt&SiO 2 nanomotors are unable to move. The same group also found that the introduction of PtO due to energetic O 2 plasma during the sputtering process leads to a reversed moving direction of the nanomotors ( Figure 8B). [140] Therefore, by controlling whether or not to prefill the sputtering chamber with inert argon gas, or treating the  Figure 8C). [37] Such difference is attributed to the role of the oxide surface enhancing the Ir catalytic performance, the catalytic efficiency of Ir&PS may be too low to enable a localized asymmetric environment. Different components have different functions, when combined in one nanomotor, they can be synergistic or cancel each other out. By introducing different functional units into nanoparticles, multimode propulsion can be achieved, enabling specific motion patterns such as "braking" and directional motion. For example, Gradilla et al. prepared three-segment Au-Ni-Au nanomotors, while being propelled by acoustic waves, the Ni segment  [54] Copyright 2021 American Chemical Society. B) Sputtering fabrication of Janus Pt&SiO 2 nanomotors and how PtO reverses the moving direction. Reproduced with permission. [135] Copyright 2022 American Chemical Society. C) Schematic of catalytic Ir/SiO 2 Janus micromotors powered by hydrazine and how changing SiO 2 into PS eliminates the motion. Reproduced with permission. [37] Copyright 2014 American Chemical Society. D) Schematic illustration of the Au&PMO/enzyme dual-propulsive Janus nanomotor. Reproduced with permission. [137] Copyright 2022 American Chemical Society. E) Schematic illustration of structure-dependent optical modulation of a TiO 2 &Au and an Au&TiO 2 nanobowl motion under the acoustic field. Reproduced with permission. [138] Copyright 2019 Wiley-VCH.
www.advancedsciencenews.com www.advintellsyst.com allowed magnetically guided motion, so that control of nanomotor's speed and direction is realized. [141] Prof. Li and Zhao's group fabricated Janus Au&PMO nanoparticles through anisotropic growth of PMO on Au nanospheres, followed by selective loading of glucose oxidase and catalase on PMO surface ( Figure 8D). [142] The Janus nanomotor thus can be propelled oppositely by thermophoresis (NIR induced photothermal effect of Au) and diffusiophoresis (tandem catalysis of glucose into H 2 O 2 and then into O 2 ). When placed in a glucose solution, the introduction of NIR produces a halt in the nanomotor's motion, exhibiting a "break"-like ability. By changing the relative positions of the different components in the nanomotor, it is also possible to create incredible changes in the motion behavior. Tang et al. fabricated nanobowls by sequentially sputtering Au and TiO 2 on PS nanospheres and then etching the PS section ( Figure 8E). [143] The nanobowls are always propelled toward the exterior side by ultrasound, which is a morphologydependent propelling process. However, under UV irradiation, redox reactions occur on Au and TiO 2 , and this propulsion force generated by localized electric field always points from Au toward TiO 2 , thus varies with the deposition sequence of the two components. When depositing Au first, so the nanobowls have Au interior and TiO 2 exterior, the nanobowl moves in accordance with ultrasound and UV. Yet when reversing the position of Au and TiO 2 , the nanobowl moves backward when switching from ultrasound to UV, thus realizing unique motion behaviors.

Effect of Surface Property on Nanomotors' Motion
The surface of nanomotors is also crucial to their motions, as surface properties determine their interactions with their surroundings. Gao et al. fabricated carbonaceous nanoflasks and controlled the surface of the nanoflasks from hydrophilic to hydrophobic by different calcination temperatures ( Figure 9A). [29] When loaded with glucose oxidase, nanoflasks move with glucose as the fuel, which is due to the glucose and gluconic acid gradient generated by the enzymatic reaction. Interestingly, they found that while the hydrophilic nanoflasks move with the opening pointing forward, the hydrophobic nanoflasks move backward. They attributed that this phenomenon is due to that the nanoflask's diffusiophoretic factors (α) of glucose and gluconic acid vary with hydrophilicity. The hydrophilic nanoflasks have a larger diffusiophoretic factor for glucose (α glu > α glu acid ), and reversed α glu acid > α glu for hydrophobic nanoflasks. As the concentration gradient is in opposite direction for glucose and gluconic acid, hydrophilic and hydrophobic nanoflasks are thus propelled in opposite directions.
Surface groups also enable specific interaction between nanomotors and functional entities. Prof. Ma's group designed and fabricated a unique nanomotor with replaceable engines. MSNs are fabricated, and the surface is modified with azobenzene groups. [144] Based on the supramolecular assembly and de-assembly between azobenzene and β-cyclodextrin (β-CD), the surface of the MSNs can be attached with Pt@β-CD, Fe 3 O 4 @β-CD and urea@β-CD "engines", and the attached "engine" can be in situ switched based on stimuli, thus achieving replacing of the engine and switching of propelling manner. Such surface group interaction-induced variable motion makes the nanomotors able to overcome a broader range of environments.

Effect of Motor Size on Nanomotors' Motion
Particle size also has a significant effect on its motion behavior, most notably the effect of size on rotational diffusion coefficient Figure 9. A) Mesoscale simulation of fluid flow around hydrophilic (blue) and hydrophobic (yellow) nanobottles, the difference in flow direction is attributed to that the two nanobottles have different diffusiophoretic factors for concentration gradient of glucose and gluconic acid. Reproduced with permission. [29] Copyright 2019 American Chemical Society. B) Different motion behavior of Pt&SiO 2 nanomotors with different sizes. Reproduced with permission. [140] www.advancedsciencenews.com www.advintellsyst.com and reorientation time. For submicron nanoparticles, their reorientation time is generally less than 0.01 s, so they are constantly rotating, and it is difficult to show directional motion. Therefore, most of the studies on submicron nanomotors are on their enhanced diffusion (enhanced Brownian motion). Only with particles larger than one micron or even five microns, where the rotation is slightly weaker and propulsion is dominant, will the study on the directional motion be possible. When the structure of the nanomotor and the ratio of each component remain the same, the size also has an effect on its movement rate. This effect is a combination of several factors such as propulsive force and resistance, so specific analysis is needed for each situation. For example, Wu et al. demonstrated that for Janus Au&polymer nanomotors with a diameter of 1, 5, 10, and 20 μm, the smaller size, the faster movement under NIR radiation. [145] They believe that while the photothermal induced temperature increase is the same, for smaller sizes, the temperature gradient is stronger across the smaller diameter.
Another key impact of size is on the generation of bubbles. It is generally accepted that for spherical nanomotors, a diameter of several microns is required for the formation of O 2 bubbles, as "nucleation" of the bubble demands a low curvature. [62] Ma et al. fabricated Pt@SiO 2 nanobowls by sputtering Pt and SiO 2 on PS nanospheres and then etching the PS ( Figure 9B). [145] They found that for nanobowls with a diameter of 500 nm, nonlinear enhanced diffusion is observed in H 2 O 2 solution. For 1.5 μm ones, the nanobowls move toward the Pt side through the electrophoretic mechanism. For larger 3 μm ones, bubbles are formed on the Pt surface, thus the nanobowls are propelled toward the SiO 2 side.

Effect of Morphology on Nanomotors' Motion
The effect of morphology on nanoparticle movements enables more precise manipulation of motion. In the vast majority of cases, changes in nanoparticle morphology do not change the mechanism of propulsion, but they do change the interaction between the nanoparticle and its surroundings, which in turn changes the speed, direction, etc., of motion.
Nanomotors of different morphologies, even with the same composition and surface properties, exhibit different motor behaviors. Baraban et al. observed that when flowing through a microfluid system with a chemical gradient, the chemotactic attraction of catalytic nanomotors varies with morphology. [146] Janus nanospheres are more sensitive than Janus nanorods and make larger "turns," which they believe is attributed to the stronger rotational diffusion of nanospheres that makes them easier attracted toward concentration gradient. Magdanz and coworkers demonstrated that for thermoresponsive "nanojets" that change configurations with varied temperatures, its propelling activity also changes ( Figure 10A). [147] Their "nanojets" are fabricated by rolling a Pt and thermoresponsive polymer multilayer films, and the rolled-up structure can bubble-propel in H 2 O 2 environment. Yet as temperature rises and the structure unfolds back into a film, bubble stops forming due to changes in surface curvature, and thus the nanostructure stopes moving. This work offers possibility for convenient control of motion during nanomotor's operations. Wilson and co-workers also demonstrated that for a stomatocytes-shape polymeric nanomotor with Pt nanocrystals loaded in the cavity, the movement can be easily changed with a minor change in morphology. [123] Normally, Pt catalyze H 2 O 2 in the solution for diffusiophoresis propelled movement. Yet, a rise in temperature leads to the collapse of nanomotor's surface polymer brushes, and "close" small opening of the stomatocytes, which induces a sudden halt of motion. Such minor morphology alteration induced motion change is critical for precise stimuli-responsive control of nanomotors.
The surface roughness also has a significant effect on the motor motion rate, as the roughness directly affects the strength of the interaction between the particles and the solution. Hormigos et al. prepared tubular micrometer motors with an inner layer of noble metal catalyst and an outer layer of carbon material with different roughness. [148] They found that for the motor with rough surface, bubble propulsion force is offset by the large friction force, resulting in a slower propulsion speed. In contrast, the motor with smooth surface has a faster propulsion speed. Similar conclusions have since been obtained using two-dimensional metal sulfides, and these works provide the basis for further understanding of the interaction with the solution during motor motion. [149] The fine-tuning of nanomotor morphology can be achieved by special synthesis methods. In certain cases, nanomotors show a gradual change of motion rate along with the morphology, which helps us to understand the relationship between morphology and motion more deeply. Ma et al. utilized an interfacial dynamic migration strategy for the synthesis of streamlined mesoporous nanotadpoles with tunable head curvatures ( Figure 10B). [121] By loading Fe 3 O 4 nanocrystals into the cavity, the nanotadpoles selfpropel in the presence of H 2 O 2 . By carefully studying the motion of different nanotadpoles, they found that curvature significantly affects the mean-square-displacement and directionality. The nanotadpoles with a curvature of %3.7 Â 10 À2 nm À1 have the fastest mobility and highest diffusion coefficient, which is attributed to a balance between fluid resistance and propelling force. Through a shielding strategy, Ji et al. fabricated a Janus Au&polymer nanomotor, in which the polymer brush length can be tuned by polymerization time. [150] They demonstrated that the grafted polymer brushes enhance the translational diffusion of the nanomotors, and the longer the brushes, the better effect. They believe that this is because grafted water-soluble polymer brushes influence the local fluid flow field and thus enhance the self-diffusiophoretic force on the nanomotors.
For some nanomotors, special architectures generate very unique motion behaviors. Wu et al. fabricated a MnO 2 dumbbell nanomotor ( Figure 10C). [151] Unlike conventional bubble propelling nanomotors where one bubble is generated at a time, the two concaves on the sides of the dumbbells can enable the simultaneously formation of a pair of bubbles. The two flapping bubbles act as "paddles" for swimming in fluid, which lead the nanomotors to move along a fluctuating circle unlike any other bubble propelling nanomotors. Xuan et al. fabricated a carbonaceous nanobottle through the emulsion templated method. [122] Upon NIR radiation, the photothermal effect of the nanobottle's carbon shell causes a rapid increase in the temperature of the water inside the nanobottle and thus the ejection of the heated fluid from the cavity. Such propulsion induced a noncontinuous www.advancedsciencenews.com www.advintellsyst.com motion that enables switching on/off the motion, which is very different from conventional solid nanoparticles. Lv et al. first fabricated Pt&PS nanoparticles through sputtering, then an Ag nanowire with controllable length is grown at the intersecting point of Pt and PS ( Figure 10D). [152] Without the Ag section, the nanospheres move away from the Pt cap in H 2 O 2 solution. Surprisingly, with the Ag nanowire, the translational motion instantly turned into a circular rotating motion with Ag "tail" pointing toward the center. This is because the addition of Ag turned the Pt-only catalysis of H 2 O 2 into an electrochemical decomposition concerning cathode (Ag) and anode (Pt), therefore forming an asymmetric distribution of protons from Ag to Pt, and thus changing the direction of propulsion force. The effect of morphology on motion is not as significant as composition, but it allows the researcher to better understand the motion mechanism of the nanomotor, while achieving precise control of motion by fine-tuning the morphologies.

Conclusion and Perspective
In summary, over the past two decades, researchers have systematically investigated the mechanisms of nanoparticle self-driving and developed a variety of nanomotors with different structures for a variety of applications. The effect of structure on the motion properties of nanomotors has also been studied in-depth, with composition, surface properties, size, and morphology as the four most significant factors most systematically studied and analyzed. A deeper understanding of the structure-motion relationship can help researchers achieve more precise control of nanomotor motion, enhance its application prospects in a variety of fields, and lay a solid foundation for the dream of nanorobots. Although fruitful results have been achieved, there is still much to focus on and explore in the field of nanomotors. How to realize the construction of nanomotors with new propulsion mechanisms. Although the synthesis technology of nanomaterials has made great progress over the years, and a series of completely new synthesis methods have been discovered; however, the current propulsion mechanism of nanomotors is relatively simple. It is usually a simple one-step reaction (e.g., catalyzing hydrogen peroxide to produce oxygen, absorbing light to produce heat, etc.) that generates a propulsive force to drive particle displacement. Unlike synthesized nanomotors, many microorganisms move in more complex ways. Bacteria, for example, undergo chemical reactions inside and outside that alter the local internal environment, driving structural changes in specific proteins that drive flagellar/ciliary oscillations for movement. [153,154] This mechanism allows microorganisms to perform relatively complex movements. The motility of microorganisms is directly linked to their Figure 10. Examples of how morphology affects nanomotor's motion. A) Schematic representation of the reversible rolling-unrolling of polymer-Pt films and images of how structure affects its bubble propelling ability. Reproduced with permission. [142] Copyright 2014 Wiley-VCH. B) TEM images with different magnifications of the Fe 3 O 4 nanocrystals loaded streamlined mesoporous silica nanomotors and schematic illustration of how curvature affects motion. Reproduced with permission. [116] Copyright 2021 American Chemical Society. C) TEM image of the dumbbell-shaped MnO 2 nanomotors and the unique dual-bubble generating motion. Reproduced with permission. [144] Copyright 2018 American Chemical Society. D) Overlaid optical micrographs of the linear movement of PS&Pt Janus nanomotors and rotation movement of PS&Pt&Ag nanomotors in 1 wt% H 2 O 2 . Reproduced with permission. [145] Copyright 2020 Wiley-VCH.
www.advancedsciencenews.com www.advintellsyst.com complex structures, such as the flagella of bacteria, which have internal structures that far exceed the complexity of current nanomaterials. Therefore, fabrication of nanomaterials with advanced nanostructures is the foundation for achieving nanomotors with more complex motion. How to explore more ways to study motion mechanisms. The current research on motion mechanism is still mainly focused on direct observation of motion behavior and theoretical simulation; however, these methods are increasingly insufficient to clarify the complex motion behavior. For example, for submicron nanoparticles, traditional optical microscopy cannot observe the particles, and dark-field microscopy can only see the translation of a bright spot, but not the specific details such as rotation. When nanomotors are anisotropic in shape, its moving direction is also unable to be observed. When the motion of a certain nanomotor involves multiple propulsion mechanisms, the occupancy of each mechanism is difficult to be studied independently. Traditional computational simulations are mainly based on static force analysis, yet nanomotors move with complex interactions with surroundings. At the same time, simple simulations often fail to take into account complex factors in the environment, such as the interference of other substances, the influence of ions and macromolecules, etc. Therefore, the introduction of newer research equipment and the development of more updated research methods will contribute to the further development of the field of nanomotors.
How can the interaction of particles with multiple interfaces in complex systems be introduced into the study of motion? Current nanomotor researchers, in the vast majority of cases, observe the motion behavior of nanoparticles in static aqueous solutions. However, in practical applications such as in vivo drug delivery and contamination treatment, nanomotors are supposed to move in a constantly flowing fluid and encounter various boundaries, such as water-oil interface in wastewater treatment, pipe walls, and biological barriers in vivo drug delivery. How will the fluid itself interfere with the motion of the nanomotor? How does the motion of the nanomotor change its interaction with various barriers? [21] There are a small number of current studies, such as the margination of particles with different morphologies and the counter-current motions, [155][156][157] but the models established in these works are still far from the practical application environment, and there is still a need to establish newer models and conduct more in-depth studies.
How to design nanomotors compatible with biological environments. Most of the current research on nanomotors is in the pure aqueous phase. However, the solution environment in practical application scenarios is very complex, including various ions and biopolymers. How will the complex solution environment affect the propulsion? How will the self-driven nanomotors (e.g., photothermal, catalytic hydrogen peroxide decomposition) change the intraorganismal environment? The design of nanomotors that are compatible with the environment of living organisms can, on the one hand, help to improve the application prospects of nanomotors. In addition, studying the effect of nanomotor motion on the biological environment can also help to develop new applications of nanomotors.
How to make the leap from a nanomotor with single propulsion mode and only simple motion behavior to a nanorobot with multiple propulsion modes and complex motion behavior? The current autonomous motion of nanoparticles is still relatively simple, basically forward moving and rotating, and the controllability is poor. To prepare nanoscale robots capable of complex motions, nanoparticles should theoretically be able to perform multiple motions such as forward and backward, lift and lower, move left and right, as well as rotate in a controlled manner. This requires multiple propulsion units with different orientations in a single nanoparticle, each of which can receive on/off commands individually and perform independent propulsion. Such a design requires, on the one hand, the development of a wider variety of methods to propel nanoparticles and, on the other hand, a combination with the controlled design and synthesis of nanomaterial structures to prepare multicomponent, complex but well controlled nanomotors. This will require relentless efforts by researchers, but although the road is arduous, the results are worth waiting for.