Light‐Activated Micromotors in Air Propelled by Thermal Convection

Abstract

Micromotors are an attractive cutting‐edge technology that exhibit controllable motion in response to chemical reactions or external stimuli. These nature‐inspired materials are widely explored for use in environmental remediation, and drug delivery, other emerging applications. Until now, the micromotors field is restricted to applications in aqueous environments, as achieving controllable motion in air while overcoming gravity remains a significant challenge. Herein, for the first time, to our knowledge, we introduce a system capable of overcoming gravity to achieve light‐induced thermal convective motion in air, driven by near‐infrared light excitation. The micromotors are composed of spiky, pollen‐like ZnO microparticles coated with gold nanoparticles, which interact photothermally with the NIR light, generating a thermal gradient that induces propulsion of the micromotor system. Lanthanide‐doped upconverting nanoparticles are deposited onto the micromotor surface to enable nanothermometric monitoring of surface temperature, providing critical information needed to describe the system's thermal behavior in air. This micromotor platform provides a versatile approach to overcome gravity and induce a controllable movement in a gaseous matrix, opening new opportunities to develop proof‐of‐concepts and applications using this aerodynamic micromotor approach.

Near‐infrared‐driven micromotors in air can overcome gravity and generate convective motion by a light‐induced thermal mechanism driven by a temperature gradient in the air caused by the interaction of the gold nanoparticles present on the surface of ZnO and NIR irradiation. NaGdF₄:Er, Yb upconverting nanoparticles enable temperature quantification on the micromotors' surface through luminescence nanothermometry. Tracking of motion and speed quantification is achieved with a microscope camera, light scattering and software tracking. Schematic created using BioRender.

1. Introduction

Light‐driven micromotors have garnered significant attention in recent decades, owing to the high spatiotemporal control enabled by using laser excitation, and the fuel‐free nature of the system.^{[ 1 ]} A variety of applications have been envisioned for these micromotors, from water remediation to drug delivery, and beyond.^{[ 2} , ³ , ^{4 ]} Controllable motion of microscale systems is especially attractive using near‐infrared (NIR) excitation, as it is abundantly available, comprising nearly half of the solar spectrum, and can be reliably transmitted through most translucent media.^{[ 5} , ^{6 ]} NIR‐driven micromotors often operate via thermally induced propulsion mechanisms, including thermophoresis, electrophoresis, and photophoresis, all of which rely on localized heating to create thermal gradients that drive fluid motion or particle displacement.^{[ 7 ]} In particular, some systems in liquid media exploit light‐induced thermal convection, where a photo‐active component absorbs light and increases in temperature, transferring heat to the surrounding fluid.^{[ 8 ]} This creates a local thermal gradient that triggers buoyancy‐driven flows, transporting or propelling micromotors without the need for chemical fuels.^{[ 9 ]} To date, such thermally activated micromotors have been largely confined to aqueous media, where the fluid not only supports the particles but also facilitates heat dissipation and convective motion.^{[ 10 ]}

However, achieving controllable motion in air presents significant challenges due to the absence of buoyant support and the dominance of gravitational forces.^{[ 11 ]} In contrast, photophoresis mechanisms have been used to move aerosols or aerodynamic microparticles in gaseous media.^{[ 12 ]} Classic photophoresis refers to the motion of particles in gases caused by anisotropic surface heating and differential interaction with surrounding gas molecules.^{[ 13 ]} While the gravitational force is similar in the mesosphere and troposphere, thermal forces can be much stronger in the mesosphere, making photophoresis more effective there. In contrast, achieving photophoresis in the troposphere is more challenging.^{[ 14 ]} To achieve tropospheric photophoresis, various mechanisms have been used to overcome gravity, including gas drag from convective flows, magnetic torques for orientation control, and experimental conditions such as microgravity to isolate thermal effects.^{[ 15 ]}

Until now, to our knowledge, all reported micromotors have been confined to aqueous or liquid environments, primarily due to the stabilization provided by the fluid medium, the difficulty of overcoming gravity in air, and the potential to interact with fuels or target compounds dissolved in the solvent environment.^{[ 16 ]} With regard to overcoming gravity in air, the size of aerosols at the microscale plays an essential role in efficient aerodynamics, as seen with natural structures such as pollen and dandelion seeds.^{[ 17 ]}

Drawing on inspiration from nature, we have introduced the first micromotors that can produce light‐activated motion in an air matrix, to our knowledge. The micromotor platform developed in this study is based on ZnO microparticles with a spiky‐pollen‐like architecture, with gold nanoparticles (Au) conjugated to the ZnO surface (Figure 1 Ai ). The light‐induced thermal convection mechanism is initiated by a thermal gradient in the air caused by the interactions among the NIR radiation, ZnO, and Au nanoparticles (Figure 1Aii ). To observe the local temperature generated by the interaction of Au with the NIR excitation, we further decorated the ZnO microparticles with NaGdF₄:Er, Yb upconverting nanoparticles (UCNPs), which enable nanoscale ratiometric luminescence thermometry. Using a combination of optical tracking, thermal imaging, nanothermometry, and computational simulations, we thoroughly characterized and described the motion of this airborne micromotor system.

Figure 1.

A) Scheme illustrating i) micromotor assembly, and ii) NIR‐light driven motion of micromotors in air by induced photophoresis. B) SEM micrographs of ZnO─Au─UCNPs micromotors with their corresponding EDS elemental mapping images for Zn, O, Au, and Gd. Scale bar is 10 µm. Schematic was created using BioRender.

2. Results and Discussion

Nitrogen adsorption–desorption isotherm analysis (Figure S1, Supporting Information) revealed that the synthesized spiky‐pollen‐like ZnO microparticles possess a specific surface area of ≈1.3948 m² g⁻¹, compared to 0.1683 m² g⁻¹ for the commercial spherical Zn precursor—an increase of roughly 88%. Transmission and scanning electronic microscopy (TEM, and SEM, respectively) images confirmed the hexagonal morphology of the NaGdF₄:Er, Yb UCNPs, faceted morphology of the Au, and spiky‐pollen‐like architecture of the ZnO, with an average size of 189.4 ± 10.7, 86.2 ± 11.1 nm, and 12.4 ± 4.2 µm, respectively. Additionally, powder X‐ray diffractograms (PXRD) indicated the crystalline nature of each particle, corresponding to their expected patterns (Figure S2, Supporting Information). Following the sequential adsorption of Au and UCNPs onto the ZnO surface via electrostatic interactions (Figure S3, Supporting Information), SEM imaging combined with energy‐dispersive spectroscopy (EDS) mapping corroborated the composition of the resulting micromotors. Characteristic elemental signals from Zn and O confirmed the ZnO backbone, while the presence of Au and Gd signals indicated successful surface loading of the Au nanoparticles and UCNPs (Figure 1B; Figure S4, Supporting Information). When Au nanoparticles were synthesized in the presence of ZnO, they remained well‐dispersed in water with a polydispersity index (PDI) of 0.28 and an average diameter of 78.8 ± 46.7 nm, as shown by dynamic light scattering (DLS) analysis (Figure S5A, Supporting Information). Upon adsorption onto the ZnO surface, they formed sparsely distributed groups of nanoparticles with a mean diameter of 240.6 ± 98.1 nm (Figure S5B,C, Supporting Information). These observations suggest that Au nanoparticles aggregate upon adsorption, forming heterogeneously distributed crystalline domains while leaving available surface area for the subsequent electrostatic attachment of UCNPs.

Additionally, inductively coupled plasma–mass spectrometry (ICP‐MS) confirmed that ZnO─Au─UCNPs contained ≈0.14 µg of Au and 82.2 µg of UCNPs per mg of ZnO, based on elemental analysis of Gd and Au, respectively (Table S1, Supporting Information). Complementary to this, zeta potential analysis was performed to evaluate the surface charge evolution during the stepwise nanoparticle assembly (Figure S6, Supporting Information). ZnO microparticles exhibited a positive charge (+21.3 ± 12.5 mV) due to surface‐exposed Zn²⁺ ions. Upon Au nanoparticle adsorption, the potential decreased to +11.8 ± 2.33 mV, likely due to partial coverage and masking of Zn²⁺ sites, as well as the negative surface charge of the Au nanoparticles (–21.4 ± 10.0 mV), attributed to residual citrate from their synthesis. Citrate‐functionalized UCNPs showed a strong negative potential (–33.2 ± 5.03 mV) from deprotonated carboxylate groups, and their subsequent grafting yielded a final zeta potential of –11.8 ± 4.88 mV for ZnO─Au─UCNPs. This systematic shift confirms sequential electrostatic assembly driven by charge complementarity.

Following surface characterization, the PXR diffractograms of ZnO─Au, and ZnO─Au─UCNPs are dominated by the intense reflections of ZnO, consistent with its microscale structure and high surface area (Figures S2 and S7, Supporting Information). Although the characteristic diffraction peaks of Au (at ≈38° and ≈44°),^{[ 18 ]} fall within regions also occupied by ZnO and residual Zn precursor signals, they are not clearly resolved in the PXRD patterns of ZnO─Au and ZnO─Au─UCNPs. This is likely due to a combination of peak overlap and the limited crystalline domain size of sparsely distributed surface‐bound Au, resulting in weak scattering intensity that is masked by the dominant ZnO peaks. In contrast, the UCNPs, although also nanoscale, form larger crystalline domains with distinguishable diffraction features. These features can be visualized by plotting the PXRD data on a logarithmic scale (Figure S8, Supporting Information), where the UCNP‐related peaks matched well with those of pure UCNPs, confirming their successful incorporation.

Remarkably, the average size, spiky‐pollen morphology and crystallinity of the ZnO remained unaffected (Figures S2 and S7, Supporting Information), which demonstrates its high stability to be used as the micromotors' backbone (ZnO) throughout the assembly process, as well as the ability of the ZnO to link nanoparticles on its surface, showing its versatility for cargo transport, and stability in the absence of aqueous media. Additional PXRD analysis confirmed that ZnO, ZnO─Au, and ZnO─Au─UCNPs retain their crystalline structure after long‐term storage in ambient air (6 months), immersion in deionized water (6 days), and incubation at high humidity (95% RH at 37 °C for 24 h), as shown in Figure S9 (Supporting Information). Their stability or structure is not water‐dependent compared to other systems, including micelles, polymersomes, and liposomes.^{[ 3} , ¹⁹ , ^{20 ]} The size of the micromotors facilitated the motion and aerodynamics under NIR‐irradiation, avoiding a possible optical trapping effect, which can be more potent for nanoparticles (Figure 1Aii ).^{[ 21 ]}

To study the airborne micromotors under ideal conditions, a closed system was constructed in order to mitigate any extraneous effects due to the surrounding environment, such as temperature and pressure changes, as well as the air currents caused by laboratory ventilation systems. To carry out this proof‐of‐concept demonstration, the closed system was assembled by joining two glass test tubes vertically with the micromotors (ZnO─Au─UCNP) or control microparticles (ZnO─UCNP) inside. An 808 nm laser beam, chosen for efficient NIR absorption by Au, is directed into the test tube perpendicular to the direction of the falling particles. By inverting the tube, the particles begin to fall by gravity and pass through this laser beam, upon which their motion is then observed (Figure S10, Supporting Information). To better illustrate the transition from micromotor behavior in water to air, we also conducted experiments with water partially filling the test tube, enabling evaluation of the microparticle response in both scenarios under identical NIR‐irradiation conditions. Confirming our hypothesis, the ZnO─Au─UCNP system exhibits motion against gravity when excited with 808 nm light, as shown in Supporting Video S1 (Supporting Information). We further demonstrate that repositioning the light source alters the direction of convective flow, enabling external guidance of particle movement (Video S2, Supporting Information). This level of directional control—achieved through remote, light‐triggered actuation—is well aligned with the classification of micromotor systems, particularly for thermally actuated designs.

Optical microscopy was used to precisely track the micromotors and characterize their motion. A microscope camera was focused on a cross‐section of the tube diameter, perpendicular to the beam irradiation (Figure S11 Ai‐ii, and B, Supporting Information). The scale was determined based on the tube's diameter (Figure S12 Ai, and Bi‐ii, Supporting Information) and the reflected NIR‐light patterns within the system, allowing for motion analysis in either a broad (Figure S12Aii , and Biii, Supporting Information) or focused region (Figure S12 Aiii, Supporting Information). Tracking and motion characterization of micromotors were studied using light scattered by the micromotors following illumination with a green LED. The use of light scattering provides a facile, non‐invasive route to micromotor tracking that is similar to the principle of the technology used for nanoparticle tracking analyzers, providing a new means to study motion and speed in air at the microscale. As shown in Figure 2Ai‐ii , we chose three in‐focus micromotors and followed their trajectory as a function of time. When the 808 nm excitation is off, the micromotors begin falling as expected due to the force of gravity (Figure 2Bi ). However, when the 808 nm laser is turned on (Figure 2Aii ) using continuous‐wave (CW) excitation (6.74 W cm⁻²), the particles are able to overcome gravity and follow an oval, counter‐clockwise trajectory (Figure 2Bii ) with a generated average speed of 906.4 ± 511.6 µm s⁻¹ (Figure 2C). This average speed accounts for the micromotors' motion at different stages, including when they move in the direction of gravity, when they begin to be suspended, and when they move against gravity, each of which is shown in Figure S13 (Supporting Information). The observed trajectory is in line with natural convection within a laminar regime (Video S1, Supporting Information).

Figure 2.

Motion behavior, tracking, and speed quantification of micromotors in air. A) Bright‐field microscopy images with a representative motion trajectory of a ZnO─Au─UCNPs micromotor from an initial (up) to a final (down) position and time i) using only scattering from the green LED without 808 nm excitation, ii) with CW 808 nm excitation (6.74 W cm^{− 2}) and a green LED light, and iii) using only the pulsed 808 nm irradiation (6.74 W cm^{− 2}, 10 Hz). B) Corresponding coordinate plots of the trajectories for i–iii) ZnO─Au─UCNPs micromotors with the previously mentioned power density conditions, and iv) ZnO microparticles as a control using the pulsed 808 nm irradiation (6.74 W cm^{− 2}, 10 Hz). C) Speed of micromotors under varying power density and frequency, and D) Speed of micromotors and controls (ZnO and ZnO─Au) without and with 808 nm irradiation, respectively. Information is derived from Video S1, and Videos S3–S6 (Supporting Information). The yellow scale bars represent 200 µm.

Using only the scattering of the 808 nm excitation, we were able to further track the micromotors' movement. The average speed increased with laser power, reaching 1150.5 ± 333.4 µm s⁻¹ at 7.47 W cm^{− 2} (CW), while maintaining the same convective trajectory (Figure 2C; Video S3, Supporting Information). Subsequently, the 808 nm excitation was pulsed (6.74 W cm⁻², 10 Hz) using the minimum possible power requirements to generate motion (Figure 2Aiii ; Video S4, Supporting Information). Under these conditions, the micromotors still overcame gravity and moved toward the entrance of the 808 nm beam (Figure 2Biii ), exhibiting an average speed of 369.1 ± 151.8 µm s⁻¹, which remained unaffected by the presence of UCNPs on the surface (Figure 2D; Videos S4 and S5, Supporting Information). In contrast, ZnO particles under 808 nm irradiation did not exhibit convective motion and continued falling due to gravity (Figure S2 Biv and 2D; Video S6, Supporting Information), similar to the behavior observed for both ZnO─Au and bare ZnO particles in the absence of 808 nm excitation (Figure 2D). These findings demonstrate that the upward motion overcoming gravity is both laser power‐dependent and laser timing‐dependent, with higher power densities resulting in increased speeds and intermittent exposure (Hz) leading to slower motion. These parameters modulate the interaction between the NIR laser and the Au on the ZnO surface, which serves as the primary mechanism driving the motion—evidenced by the absence of motion when Au is not present or the laser is off (Figure 2C,D).

To better understand the transition from aqueous to airborne operation, we performed comparative experiments using the same tube setup partially filled with water. In water, the higher viscosity slows the falling ZnO particles, allowing for prolonged interaction with the NIR beam. Under these conditions, ZnO microparticles exhibited NIR‐induced motion with an average speed of 58.7 ± 26.1 µm s⁻¹, consistent with prior reports of ZnO‐based micromotors in aqueous environments, which typically exhibit speeds below 70 µm s⁻¹ under similar excitation conditions.^{[ 10 ]} When Au was incorporated onto the ZnO surface, the motion speed in water significantly increased to 430 ± 99.6 µm s⁻¹—over 7.4 times faster—demonstrating the crucial role of Au in enhancing photothermal conversion (Videos S7 and S8; Figure S14, Supporting Information). These results highlight that while ZnO can move in water via light‐induced interactions, this mechanism alone is insufficient to enable motion in air, where the presence of Au nanoparticles is needed to overcome gravity. Notably, recent reports have demonstrated gravity‐defying micromotors in water using combined thermophoretic, electrophoretic, or enhanced thermoplasmonic propulsion strategies via light‐induced.^{[ 22} , ^{23 ]} In contrast, our system demonstrates that light‐induced convection alone—without chemical fuels or electric fields—is sufficient to produce directional or upward motion in both water and air. The use of a single propulsion mechanism across two distinct media, validated experimentally, highlights the versatility and fuel‐free nature of our approach. Furthermore, the micromotors can move in the air for at least 3 min (Video S9, Supporting Information), demonstrating their potential for prolonged operation in applications where sustained action is fundamental, such as catalysis, aerosol toxin detection, and pollution control.

Given that the micromotor system is proposed to operate via light‐induced temperature differentials in the microparticle and surrounding environment, we set out to quantify the changes in temperature induced under NIR excitation at 808 nm both in the air and at the surface of the micromotors/particles. Due to technical restrictions, we chose to restrict our focus to studying the particles' temperature and spatiotemporal properties in a 2D plane perpendicular to the light beam (Figure S15, Supporting Information) in order to enable acquisition of thermal images in a cross‐section of the system (Figure S16, Supporting Information). In all cases, we studied each component of the micromotor system: ZnO alone, ZnO─UCNPs, ZnO─Au, and ZnO─Au─UCNPs with and without 808 nm (exciting Au). The properties of the 808 nm excitation source (Table S2, Supporting Information) were chosen such that they would not affect the thermal equilibrium of the system in the absence of Au. Thermal imaging confirmed there was no change in temperature of the air for the ZnO scenario (Figure S17, Supporting Information). Thus, the observed temperature increase of the air when ZnO─Au particles interact with the 808 nm light (Figure S18, Supporting Information) is required to achieve motion (Figure S19, Supporting Information). A similar experimental setup was applied for the aqueous case, using ZnO and ZnO─Au microparticles. Thermal imaging revealed a notable increase in water temperature with ZnO─Au under 808 nm irradiation, confirming that local heating can also induce convective motion in liquids (Figure S20, Supporting Information). This supports the hypothesis that thermally induced convection, initiated by the Au–NIR interaction, plays a central role in the propulsion mechanism, both in air and in water.

To further corroborate this finding, we used upconversion nanothermometry to observe the local change in temperature at the surface of ZnO microparticles in the presence and absence of Au, using upconverted green luminescence from Er³⁺ in the UCNPs under 976 nm excitation. The intensity ratio of the ²H_11/2→⁴I_15/2 and ⁴S_3/2→⁴I_15/2 transitions of Er³⁺ follows a Boltzmann distribution (Equation S1, Supporting Information) and is well‐established for luminescence thermometry (Figure 3 Ai ).^{[ 24 ]} A calibration curve was established (Figure S21A, and Equation S2, Supporting Information), which enabled ratiometric temperature readouts at the surface of ZnO─UCNPs and ZnO─Au─UCNPs with and without 808 nm excitation (Figure S21B, Supporting Information). For ZnO─UCNPs and ZnO─Au─UCNPs, temperature increases of 2.8 and 12 °C were observed (Figure 3Aii ), respectively (relative to room temperature), confirming the photothermal effect generated when Au absorbs 808 nm light (Figure S22, Supporting Information). The comparatively larger temperature differential observed upon interaction of Au with 808 nm light was also observable by thermal imaging, confirming successful heat dissipation to the surrounding environment (air). Notably, thermal imaging shows a temperature increase in the region closest to the entrance of the 808 nm beam. This is likely due to scattering of the beam upon interaction with the curved surface of the glass, generating a region of higher power density inside the tube closer to the introduction of the beam and a lower power density as the beam progresses laterally through the closed system. This power density discrepancy of the 808 nm irradiation, caused by the geometry of the laser beam, (Table S2; Figure S23A–C, Supporting Information) interacts anisotropically with the micromotors, which are falling in parallel through the diameter of the tube, generating a proportional decrease in the air and water temperature due reduced temperature differentials between the micromotors and surrounding environment. Despite the low thermal conductivity of the air, which restricts efficient thermal dissipation, a 1.5 °C increase in the air temperature was observed for the scenario with ZnO─Au micromotors (Figure 3Bi ). This explains why the micromotors exhibit gravity‐defying convective motion, moving upward in the region where the temperature increase is observed at the entrance of the 808 nm beam into the system. Similarly, the observed temperature increases of 2.8 °C for ZnO and 7.7 °C for ZnO─Au microparticles in water correlate with the convective motion behaviors previously discussed, with the higher temperature gradient in the ZnO─Au system corresponding to significantly faster movement. To further describe the motion of the micromotors, the Grashof (Gr) number,^{[ 25 ]} which is used to determine the relative strength of buoyancy and viscous forces in a fluid flow,^{[ 26 ]} was calculated, as it has been employed in similar closed systems to evaluate if a fluid presents a natural convection by photophoretic effects (Equation S3, Supporting Information).^{[ 8 ]} Gr was determined by using the previously quantified difference of temperatures of the air regions, the 16 mm diameter of the tubes, and the specific parameters of the air, and the water (Table S3, Supporting Information). The calculated values of the Gr range between 10³ and 10⁶ and are characteristic of the laminar boundary layer and natural convection in air (Figure 3Bii ).^{[ 27} , ^{28 ]} In the aqueous system, the calculated Grashof numbers exceeded 2400 for both ZnO and ZnO─Au systems (Figure S24A, Supporting Information), indicating the presence of natural convection consistent with the observed micromotor motion.^{[ 29 ]} Those indicates that the light‐induced thermal convection mechanism is responsible for the air, and water convection, which describes an intrinsic difference in the mass density of the air, water, and micromotors, as well as the thermal difference.^{[ 13} , ^{30 ]}

Figure 3.

Motion mechanism determination of the micromotors in air under 808 nm excitation. A, i) Upconversion emission spectra obtained from ZnO─UCNPs, and ZnO─Au─UCNPs micromotors (λex: 808, and 976 nm, 10 Hz), and A, ii) quantified surface temperature of ZnO─Au─UCNPs micromotors, and Zn─UCNPs microparticles under 808 nm irradiation. B, i) thermal photograph of the air contained in a glass tube upon 808 nm irradiation of ZnO─Au─UCNPs (0.55 W, 10 Hz). B, ii) The change in temperature (ΔT) and Grashof number of the air observed as a function of NIR exposure time under 808 nm excitation. Temperature scale in °C. C, i) Thermal behavior of a region of micromotors, and the air. C, ii) Surface velocity (µm/s) and direction of the air flux (arrows) as a function of the temperature distribution. Simulated using Comsol 6.3.

Computational simulations were used to provide a precise description of the micromotor system and surrounding environment based on the calculated Gr. We initially estimated different temperature points along the tube diameter, which depend on the generated heat of the ZnO─Au micromotors under different 808 nm power densities (Figure S23C; Table S2, Supporting Information). The simulation parameters were based on our experimental conditions, as well as the temperatures evaluated using nanothermometry and thermal imaging. The simulated thermal heat dissipation matched well with the experimental results, where the higher‐temperature micromotors dissipated heat into the air, despite its low thermal conductivity (Figure 3Ci ). The simulations predicted natural air convection with the motion toward the laser beam source, described by a laminar layer. The oval‐like counter‐clockwise motion observed occurs due to the influence of the stable adiabatic boundary and the colder air adjacent to the air within the laser beam path. The simulated air velocities were estimated ≈1500 ± 500 µm s⁻¹ depending on the region of the motion path being studied, in agreement with the changes in speed of the micromotors as observed experimentally at different regions of the motion paths (Figure 3Cii ; Figure S12 , Supporting Information). Comparable simulation results were obtained for the water‐based system, confirming similar thermally induced convective behavior under 808 nm irradiation (Figure S24B,C, Supporting Information). As mentioned previously, the experimentally tracked speeds and trajectories are in good agreement with the simulated values, underscoring the validity of our strategy for achieving photophoretic motion of micromotors in air for the first time.

The water‐based simulation revealed the same motion dynamics observed experimentally, though with slower velocities due to water's higher viscosity. These findings reinforce the critical distinction between the two environments: in water, motion is facilitated by enhanced thermal coupling and suspension but is hindered by viscous drag; in contrast, the transition to air overcomes gravitational constraints through light‐induced thermal convection and benefits from reduced resistance, resulting in enhanced kinetic behavior due to the favorable aerodynamic conditions. Furthermore, directional control of the system was simulated by repositioning the beam on the opposite side of the chamber. The resulting thermal profile exhibited a similar temperature gradient (Figure S25A, Supporting Information), while the simulated air velocity fields showed consistent convective flow reorientation (Figure S25B–D, Supporting Information). These results confirm that the mechanism of light‐induced thermal convection remains effective upon directional switching, demonstrating the robustness and reconfigurability of the propulsion system.

To further explore the scalability and applicability of this system in air, we simulated convection behavior in tubes of varying diameters—12, 16, and 20 mm—corresponding to a range of Grashof numbers (Equation S3, Supporting Information). The results revealed consistent convective flow patterns and velocities across this range, indicating that the propulsion mechanism is robust within practical geometries (Figures S26A–C, Supporting Information). These findings suggest that the micromotor platform could be implemented in semi‐confined environments with diameters ranging from 4 to 35 mm, supporting its potential for use in real‐world systems.

3. Conclusion

For the first time, to our knowledge, we present a micromotor system with sustained motion in an air matrix. The designed system consisted of spiky‐pollen‐like ZnO microparticles coated with Au nanoparticles and NaGdF₄:Er,Yb UCNPs. The micromotors show aerodynamic features, enabling the micromotors to overcome gravity by generating natural thermal convection via light‐induced thermal convection. We used upconversion nanothermometry to evaluate the local temperature at the surface of the micromotors, and thermal imaging to evaluate the temperature of the surrounding air, confirming that thermal dissipation from the micromotors to the air facilitated photophoretic motion. All of our experimental results were corroborated with simulations, enabling a precise description of the mechanics of the system. Importantly, comparative experiments with water‐based systems allowed us to demonstrate the transition from aqueous to airborne operation. While convective motion was observed in both cases, propulsion was significantly more efficient in air due to enhanced light fluence and reduced viscous drag, which was achievable only once gravitational constraints were overcome. This breakthrough in the micromotors field paves the way for new, controllable microscale technologies that operate in an air matrix. Notably, these micromotors can transport other types of nanoparticles besides UCNPs, enabling a variety of different applications depending upon the properties of the nanoparticle cargo. We envision a variety of applications, including gas‐phase heterogeneous catalysis, aerosol toxin detection, and pollution control, operating in channels, cartridges, or tubing within a practical diameter range compatible with natural convection, as supported by our simulations. This demonstrates the potential of this aerodynamic micromotor platform to impact environmental and catalytic processes. Given the light‐driven, fuel‐free nature of this system, future development could also explore solar‐responsive micromotors for passive environmental remediation in open or semi‐confined airspaces.

4. Experimental Section

Synthesis of ZnO Microparticles

ZnO microparticles were synthesized using the conventional method employing the Kirkendall effect.^{[ 18 ]} 0.65 g (10 mmol) of spherical metallic Zn powder was added to 25 mL of deionized H₂O under stirring. 0.75 g of Zn (NO₃)₂.6H₂O was added to 5 mL of 0.5 M NaOH water solution. After homogenization by continued stirring, the zinc nitrate solution was added to the Zn powder dispersion. The reaction was carried out in a Teflon‐lined stainless steel autoclave at 220 °C for 2 h. The product was washed once with water, followed by 2 times with ethanol, ensuring the purity of the final product. The product was vacuum‐dried at 40 °C at 20 mbar for 25 min.

Assembly of ZnO─Au Micromotors

Au nanoparticles were synthesized in situ during the adsorption process on the ZnO surface, based on a previously published method.^{[ 18 ]} Briefly, 0.1 g of ZnO microparticles were dispersed in 20 mL of deionized water under constant stirring at 800 rpm. Then, 1 mL of 0.01 m sodium citrate and 1 mL of 0.01 M HAuCl₄ were added to the ZnO dispersion. Subsequently, 2 mL of 0.01 m ascorbic acid were added dropwise under continuous stirring, and the reaction was allowed to proceed for 5 min. A visible color change from transparent to purple indicated the formation of Au nanoparticles on the ZnO surface. The resulting ZnO─Au micromotors were separated by decantation and washed twice with deionized water and once with ethanol to remove excess unbound Au nanoparticles. The final product was vacuum‐dried at 40 °C and 20 mbar for 25 min.

For characterization of Au nanoparticles alone, including TEM imaging, hydrodynamic size (DLS), and zeta potential, a similar synthesis protocol was followed under the same conditions but in the absence of ZnO microparticles.

NaGdF₄:Er³⁺, Yb³⁺ Upconverting Nanoparticles Synthesis

NaGdF₄:Er³⁺,Yb³⁺ upconverting nanoparticles (UCNPs) were synthesized using our previously published double‐injection thermal decomposition method.^{[ 31 ]} A mixture of 1.25 mmol of lanthanide oxides (78% Gd₂O₃, 20% Yb₂O₃, and 2% Er₂O₃) was dissolved in 10 mL of CF₃COOH/water (1:1 v/v) and dried at 60 °C. Meanwhile, in a separate 100 mL three‐neck round‐bottom flask, 12.5 mL of oleic acid and 12.5 mL of 1‐octadecene were degassed at 120 °C for 30 min. Separately, the dried precursors and 2.5 mmol of CF₃COONa (NaTFA) were combined with 7.5 mL each of oleic acid and 1‐octadecene, then degassed under identical conditions while stirring at 300 rpm. The solvent mixture was gradually heated to 310 at 10 °C min⁻¹. The precursor solution was then injected at 1.5 mL min⁻¹, and the reaction proceeded for 1 h with continuous stirring at 300 rpm, under argon. The core/shell nanoparticles were purified by centrifugation at 4000 rpm for 15 min, followed by three washing cycles with a hexane/ethanol mixture (50 mL, 1:5 v/v). The purified nanoparticles were then precipitated and stored as a pellet in ethanol.

Citric Acid‐Coated NaGdF₄:Er³⁺,Yb³⁺ UCNPs

UCNPs' surface was modified through a ligand exchange with citrate according to a previously established method.^{[ 32 ]} 25 mg of UCNPs were initially dispersed in 5 mL of hexane. Then, 5 mL of 0.5 m sodium citrate (tribasic) solution was added to the dispersion, resulting in phase separation. The mixture was stirred for 4 h at 1000 rpm, allowing the UCNPs to migrate to the aqueous phase, indicating the removal of the oleate coating. After discarding the organic phase, 5 mL of ethanol was added to the aqueous dispersion, and then the dispersion was washed with water by centrifugation at 4000 rpm for 5 min. Finally, the citrate‐coated UCNPs were stored as pellets in ethanol.

Assembly of ZnO─Au─UCNPs Micromotors

The ZnO surface was functionalized with citrate‐coated UCNPs via electrostatic interaction with the carboxylic acid groups of citrate.^{[ 33 ]} To facilitate this, 25 mg citrate‐coated UCNPs were dispersed in 20 mL of water, and added 0.1 g of ZnO─Au micromotors. The mixture was stirred for 5 min at 1000 rpm. The micromotors were then separated by decantation after incubation in an ice bath. Then, the precipitate was washed twice with water and once with ethanol. Finally, the ZnO─Au─UCNPs micromotors were vacuum‐dried in a rotovap at 40 °C and 20 mbar for 30 min.

Light‐Induced Thermal Convective Motion Experiments, Speed Quantification, and Supporting Videos

For the thermal convective motion experiments, 0.1 g of micromotors or control microparticles were introduced into a glass tube system, similar to the thermal behavior tests for motion characterization in air. For comparison with the system in liquid, 5 mL of water were added to the glass tube containing the microparticles, creating a confined aqueous environment for evaluating convective behavior. A 6W dual light‐emitting diode (LED) microscope was used to visualize the falling micromotors and control particles. During descent, the particles were irradiated perpendicularly with 808 nm light. To reverse the direction of air convection, the opposite side of the system was irradiated with 808 nm light while the initial side was turned off. Their trajectories were recorded using a liquid‐crystal display (LCD) digital microscope camera equipped with 144 LED lights and a 180X high‐definition multimedia interface (Figure S11, Supporting Information). The microscope camera was operated with Image View software to capture and export the videos. The videos were processed using the TrackMate plugin in Image J to analyze the motion (courtesy of the National Institutes of Health). The micromotor and control trajectories were plotted as 2D coordinates, and displacement was analyzed over a 15 s duration for each test (n = 10). Average speeds were calculated based on the measured trajectory distances and the corresponding time intervals, using the video frame rate as a reference.

After falling freely along the gravity vector, ZnO─Au─UCNP micromotors demonstrated directional control, overcoming gravity and moving toward the 808 nm irradiation. At the boundary of the irradiation region, an oval trajectory emerged, indicating the onset of convective motion (Video S1, Supporting Information). Reversal of the convective flow and corresponding micromotor motion was demonstrated by irradiating the opposite side of the system with 808 nm light, confirming light‐guided directional control (Video S2, Supporting Information). By increasing the laser power of the 808 nm irradiation led to a higher speed of convective motion of the micromotors (Video S3, Supporting Information). The same directed motion overcoming gravity was observed when the experimental area was restricted to only the 808 nm‐irradiated region (Video S4, Supporting Information). Similarly, ZnO─Au micromotors exhibited comparable motion behavior under 808 nm irradiation (Video S5, Supporting Information). In contrast, ZnO microparticles exposed to 808 nm light continued their descent along the gravity vector without exhibiting motion against gravity (Video S6, Supporting Information). Motion of the ZnO and ZnO─Au microparticles in water was also evaluated under 808 nm irradiation (Videos S7 and S8, Supporting Information), allowing comparison between air and liquid environments. Furthermore, micromotors can move in air for more than 60 s (Video S9, Supporting Information).

Upconversion Nanothermometry

A luminescence intensity‐based ratiometric nanothermometry method was used based on our previously published method.^{[ 24 ]} First, the ZnO─Au─UCNPs micromotors were placed inside a capillary and exposed to 976 nm irradiation. The emission spectra were evaluated at different temperatures on a thermally‐regulated sample stage. The ln(I₅₂₅/I₅₄₅) of the upconversion emission from the ²H_11/2→⁴I_15/2 and ⁴S_3/2→⁴I_15/2 integrated intensities, and the inverse of the temperatures was plotted to determine the calibration curve, based on the thermal equilibrium described by the Boltzmann factor (Equation S1, Supporting Information). The corresponding linear fit Equation S2 (Supporting Information), with a correlation coefficient (R ²) of 0.9981, was then deduced from the calibration curve.

Then, the ZnO─Au─UCNPs micromotors and ZnO‐UCNPs control microparticles were excited at room temperature using 976 nm and/or 808 nm laser sources, and their upconverted emissions were recorded with a Princeton Instruments FERGIE BRXVR UV−NIR spectrograph with a 250 grooves/mm grating blazed at 550 nm and a 50 µm entrance slit. A 600 µm Ocean Optics optical fiber was used for coupling, along with a Thorlabs 400−750 nm (FESH0750) visible bandpass filter to isolate the green upconverted emissions. Spectra were processed using Light Field software. The temperatures were determined depending on the ²H_11/2→⁴I_15/2 and ⁴S_3/2→⁴I_15/2 integrated intensities and using Equation S2 (Supporting Information).

Thermal Measurement Acquisition of NIR‐Irradiated Air Area for Grashof Determination

A hand‐held thermography camera (HKMicro B10) was used to capture thermal images of the temperature distribution in air and water containing micromotors and control particles under 808 nm excitation. This enabled the quantification of the temperature change (ΔT) and, consequently, the determination of the Grashof number (Equation S3, Supporting Information) by using the parameters of the air and the glass tube diameter (Table S3, Supporting Information). A total of 0.1 g of micromotors or controls were introduced into a glass tube system. The system was irradiated perpendicularly using an 808 nm diode laser (6.74 W cm⁻², 10Hz) while the particles were free‐falling. The thermal camera was positioned perpendicular to the NIR irradiation (Figure S16, Supporting Information) to record thermographs at 0, 15, 30, and 180 s in order to observe temperature fluctuations in the air.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Supplemental Video 1

Supplemental Video 2

Supplemental Video 3

Supplemental Video 4

Supplemental Video 5

Supplemental Video 6

Supplemental Video 7

Supplemental Video 8

Supplemental Video 9

Mena‐Giraldo P., Mandl G. A., Quezada‐Novoa V., et al. “Light‐Activated Micromotors in Air Propelled by Thermal Convection.” Adv. Mater. 38, no. 1 (2026): e05959. 10.1002/adma.202505959

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.