Go-and-Stop Nanovehicles Enabled by Counterbalanced NIR Optical Engines

Abstract

Nanomotors are promising for diverse applications due to their ability to convert energy into directional motion. However, precise control of their motion is still difficult because creating required asymmetric nanostructures presents significant synthetic challenges. Inspired by automotive engine control, this study reports a general strategy for constructing optical nanomotors with programmable motion. For the first time, this approach implements a “Go and Stop” concept, leveraging dissimilar near-infrared (NIR) responsivities of upconversion nanoparticle (UCNP) to accelerate (“Go”) or brake (“Stop”) motion on demand. Specifically, the nanovehicle is designed via interface-guided anisotropic growth strategy, where half of a single upconversion nanoparticle (UCNP) is coated with a hydrophilic silica shell containing Au nanoseeds, while the other is covered by a hydrophobic periodic mesoporous organosilica (PMO) loaded with Roussin’s black salt (RBS). This dual hydrophilic/hydrophobic design is also applicable to other optical materials, thus demonstrating great flexibility. When excited at 980 nm, blue light emitted by the UCNP activates the RBS to release nitric oxide (NO), propelling the nanomotor forward. Conversely, light at 808 nm induces gold-mediated thermophoresis, which reduces and eventually stops NO-driven motion. Such “pull–push” effect enables the nanomotor to start, accelerate, decelerate, and stop simply by switching the excitation light. As a proof of concept, the nanovehicle is further used to noninvasively regulate cellular endocytosis and has demonstrated great potentials for programmable tumor treatment.

1. Introduction

Nanomotors are nanoscale structures that enable to transform energy into directional motions, exhibiting unique advantages over conventional nanoparticles that move randomly due to Brownian motion. − By transforming external stimuli or local environmental energy into mechanical force, nanomotors can achieve autonomous movement for various applications in drug delivery, biosensing, environmental remediation, , tumor therapy, etc. One common approach to achieve propulsion in nanomotor is via decomposition of some intracellular chemical or biological fuel molecules, such as H₂O₂, urea, , and glucose, which may release gas bubbles as the driving force. However, the gradual depletion of these fuels inevitably results in a loss of motility, thus restraining their long-term operation. Therefore, various nanomotors driven by external energy sources, e.g., magnetic fields, , ultrasound, , and light, , have been developed to solve these issues, where the nanomotor motion can be controlled by fine-tuning the input intensity or frequency of applied physical stimuli. Despite these great achievements, most of the nanomotors are still limited to unimodal propulsion, which significantly limits their applications in more complicated scenarios where precise control of velocity and distance and multidirectional movement is required.

Recently, a variety of dual-mode driven nanomotors have been reported by combining different external fields, such as magnetic field/ultrasound, magnetic field/chemical energy, ultrasound/chemical energy, ultrasound/light, and chemical energy/light. , For example, Ren et al. developed a bimetallic micromotor system that synergistically harnessed acoustic force and chemical fuel (Pt/H₂O₂) energy for bidirectional rheotaxis (>20 μm/s), where ultrasound-guided directional control enabled programmable navigation with positive or negative flow alignment by adjusting the tilt angle from 0° to 18°. In another study, Wang et al. reported dual-force enhanced biodegradable stomatocyte nanomotors powered by enzymatic H₂O₂ generation and NIR-induced plasmonic heating. Due to the combinational effect, the rapid propulsion speed (up to 96 μm/s) was achieved.

Obviously, the presence of dual propulsion significantly enhances the motion stability of the nanomotor compared to its single-mode counterparts. Moreover, the dual-propulsion driving mechanism also provides enhanced flexibility in controlling movement on demand, which is hardly achieved in single-mode propulsion systems. Nevertheless, the operation of dual-propulsion nanomotors relies on combining multiple external stimuli, e.g., magnetic field, ultrasound, chemical energy, and light, which also increases significant complexity in system design and motion control. Among the external stimuli, light excitation presents unique advantages such as ease of operation, facile frequency/intensity tuning, adjustable directionality, and minimal invasiveness. Despite these alluring benefits, there have been few studies on dual-propulsion-driven nanomotors that solely rely on light sources. This is primarily due to considerable challenges in tailoring the surface chemistry of the same nanoparticle and endowing it with dissimilar optical responses to varied light excitations.

In light of these considerations, this study develops an interface-mediated asymmetric growth strategy to construct a “vehicle-like” nanomotor, which is capable of starting, accelerating, decelerating, and stopping only driven by varying excitation lights. The nanovehicle is a specially designed upconversion nanoparticle (UCNP, NaYF₄:Yb,Tm@NaYF₄), partially covered by a hydrophilic silica shell embedded by Au nanoseeds, with the remaining portion encapsulated by a hydrophobic periodic mesoporous organosilica (PMO) loaded with Roussin’s black salt (RBS). Moreover, the proposed synthesis strategy was also used to construct other asymmetric optical nanomotors, e.g. persistent luminescence nanoparticles (ZnGa₂O₄/Mn²⁺), demonstrating great flexibility in adjusting luminescence profiles, morphology, and size. Taking advantages of dual optical properties of the nanovehicle, two near-infrared (NIR) light sources empower precise control of particle motion by serving as dual-balancing optical engines for acceleration and deceleration. Specifically, the 980 nm light enables to initiate and accelerate the nanovehicle via NO gas release from RBS salt; while the 808 nm light brakes and stops it through thermophoresis induced by Au nanoseeds (Scheme ). Furthermore, such precise control over the particle had been shown to enable modulation of cell–particle interactions, resulting in noninvasive regulation of cellular endocytosis through distinct excitation schemes. Besides, the independent mechanisms of NO-mediated cellular dysfunction and Au-induced photothermal apoptosis also allowed for programmed therapy and led to significantly enhanced synergistic effects for tumor suppression, which demonstrated the as-developed nanovehicle for promising applications in targeted drug delivery and precision phototherapy.

1. Schematic Illustration of Light-Regulated Endocytosis and Phototherapy Using Dual-Light Controlled Nanovehicles.

2. Results and Discussion

2.1. Construction and Characterization of UCNPs@SiO₂@Au@PMO-RBS Nanovehicles

In this study, the “go-and-stop” nanovehicle was developed using an interface-guided anisotropic growth approach, where one side of a light-harvesting UCNP core is covered by gold nanoparticles and the other side with periodic mesoporous organosilica (PMO) (Figure A). Experimentally, core–shell-structured UCNPs (NaYF₄:Yb,Tm@NaYF₄) were first synthesized via a conventional high-temperature coprecipitation method and subsequently coated with a dense silica layer to form UCNPs@SiO₂. Transmission electron microscopy (TEM) imaging (Figure B) confirms the uniform morphology of UCNPs@SiO₂ with an average size (∼100 nm). Notably, the SiO₂ layer thickness can be easily tuned by controlling the volume of the silicon source (TEOS) (Figure S1) and the reaction time (Figure S2). To introduce photothermal functionality, gold nanoparticles (Au seeds, ∼5 nm) were then anchored onto the surface of amino-functionalized UCNPs@SiO₂–NH₂ via electrostatic interactions to form UCNPs@SiO₂@Au (Figure C). It was also found that the distribution uniformity of Au seed on the SiO₂ layer was determined by its size (Figures S3 and S4), while the surface density was dependent on the Au-to-UCNPs@SiO₂ ratio (Figure S5). Subsequently, periodic mesoporous organosilica (PMO) was asymmetrically grown on one side of UCNPs@SiO₂@Au using an anisotropic island nucleation and growth strategy, forming UCNPs@SiO₂@Au@PMO (Figure D). The asymmetric coating ratio of PMO could also be precisely controlled by changing the addition amount of the precursor BTEE (Figure S6). Specifically, Figure E demonstrates the morphology of a single UCNPs@SiO₂@Au@PMO Janus nanoparticle, while Figure F presents the HRTEM image corresponding to the red dashed box in Figure E. Two types of lattice fringes are clearly observed in Figure F: one with an interplanar spacing of 0.31 nm corresponding to the (110) plane β-NaYF₄ and the other with a spacing of 0.23 nm corresponding to the (111) plane Au. This can also be verified by the diffraction patterns via fast Fourier transform (FFT) analysis (Figure G), where the diffraction spots can be well indexed to β-NaYF₄ (blue) and Au crystal planes (yellow). Meanwhile, the anisotropic structure of UCNPs@SiO₂@Au@PMO nanoparticle was further confirmed using high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), as shown in Figure H. The bright regions correspond to the Au seeds surrounding the UCNPs@SiO₂ particle, while the dark region indicates the presence of asymmetrically deposited PMO domains. These results are consistent with elemental mappings (Figure I–L), which indicate the spatial distribution of Si, Yb, F, and Au in their respective regions. Additional structural and compositional characterizations of the UCNPs@SiO₂@Au@PMO have also been performed. For example, Figure S7 shows the X-ray diffraction patterns of the synthesized UCNPs, UCNPs@SiO₂, UCNPs@SiO₂@Au, and UCNPs@SiO₂@Au@PMO nanoparticles, which all clearly display the characteristic diffraction peaks of hexagonal-phase NaYF₄, SiO₂, and Au. Nitrogen adsorption–desorption analysis also reveals a high-surface area of 584.13 m²/g with an average pore size of 2.71 nm (Figure S8). More importantly, this synthesis strategy enables the integration of other anisotropic nanoparticles (e.g., ZGM, ZnGa₂O₄:Mn²⁺) to fabricate persistent luminescence-based nanomotors (ZGM@SiO₂@Au@PMO), as characterized in Figures S9–S11.

1.

Synthesis and characterization of light-controlled asymmetric nanoparticles. (A) Schematic of the synthesis process of UCNPs@SiO₂@Au@PMO nanoparticles. TEM image of (B) the UCNPs@SiO₂, (C) UCNPs@SiO₂@Au, and (D) UCNPs@SiO₂@Au@PMO nanoparticles. (E) TEM image of a single UCNPs@SiO₂@Au@PMO nanoparticle. (F) High-resolution TEM image and (G) diffraction patterns of the area marked by the red dashed square in (E). (H) HAADF-STEM image of the single UCNPs@SiO₂@Au@PMO nanoparticle and its corresponding elemental mapping for Si (I), Yb (J), F (K), and Au (L). (M) Upconversion luminescence spectra of UCNPs@mSiO₂ under excitation at 980 nm light (blue line), and the absorption spectra of RBS (cyan dashed line) and Au (yellow dashed line). (N) Upconversion mechanism for blue emission excited under 980 nm light excitation.

Regarding the anisotropic growth mechanism of PMO, we propose that this morphology is attributed to differences in surface tension and specific island nucleation processes. First, tetraethyl orthosilicate (TEOS) was used as a precursor for synthesis of the UCNPs@SiO₂ domain. Then, the cationic surfactant CTAB was used as a structure-directing agent, which played crucial roles for the anisotropic deposition of PMO through hydrolysis of BTEE (organic silane [bis­(triethoxysilyl)­ethane]. Specifically, the change in total surface energy Δσ for the deposition process can be expressed as

$$\mathrm{\Delta }\sigma ={\sigma }_{\mathrm{P}\mathrm{M}\mathrm{O}‐\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}}+{\sigma }_{\mathrm{P}\mathrm{M}\mathrm{O}‐{\mathrm{S}\mathrm{i}\mathrm{O}}_2}-{\sigma }_{{\mathrm{S}\mathrm{i}\mathrm{O}}_2‐\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}}$$ 1

where σ_{PMO–solvent} and ${\sigma }_{{\mathrm{S}\mathrm{i}\mathrm{O}}_2‐\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}}$ are the surface energies of PMO and SiO₂ in the solvent, respectively; ${\sigma }_{\mathrm{P}\mathrm{M}\mathrm{O}‐{\mathrm{S}\mathrm{i}\mathrm{O}}_2}$ is the interfacial energy between PMO and SiO₂. As shown in Figure S12A, when pure H₂O is used as the solvent, the high surface tension of water results in σ_{PMO–solvent} > ${\sigma }_{{\mathrm{S}\mathrm{i}\mathrm{O}}_2‐\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}}$ , leading to Δσ > 0. It should be noted that the interfacial energy, ${\sigma }_{\mathrm{P}\mathrm{M}\mathrm{O}‐{\mathrm{S}\mathrm{i}\mathrm{O}}_2}$ , is rather low due to similar Si–O–Si backbones of both PMO and SiO₂. Thus, this means that the uniform and continuous formation of CTAB/silicate micelle coating is unfavorable under such conditions. Instead, PMO tends to grow anisotropically through hydrolysis of BTEE on the SiO₂ surface to form the Janus-type morphology (Figures S12B). From a large-view examination in Figure S13, the yield of asymmetric Janus particle can reach around 90% (green).

To endow nanovehicle (UCNPs@SiO₂@Au@PMO-RBS) fuel molecules, the Roussin’s black salt (nitric oxide donor) was loaded into the porous channels of the PMO nanocomposite. Fourier transform infrared (FTIR) spectroscopy was employed to characterize the RBS-loaded nanocomposite. As demonstrated in Figure S14, the band at approximately 1104.1 cm^–1 is attributed to the Si–O–Si stretching vibration of SiO₂; the peaks at 1655.1 and 1397.2 cm^–1 correspond to the N–O bonds of RBS; and the peaks at 2925.9, 1737.5, and 1272.8 cm^–1 are assigned to the stretching vibrations of C–H, CO, and Si–C bonds of the organic mesoporous silica (PMO). The loading efficiency of RBS molecules was also investigated. Experimentally, the UV–vis absorption spectra of RBS salt at different concentrations were first measured (Figure S15A). Then, a calibration curve of absorbance at 365 nm versus concentration was established (Figure S15B), from which the maximum loading efficiency of RBS in the UCNPs@SiO₂@Au@PMO was determined to be of 6.9 wt %. As for the study on the release of RBS, the UCNPs@SiO₂@Au@PMO-RBS was first soaked in the cell culture medium (DMEM) for 24 h. Then, the supernatants after centrifugation were collected. Next, the UV–vis absorption spectrum of the supernatants was measured, as shown in Figure S15C. It clearly shows there is no significant absorbance changes in the solution of DMEM, suggesting that RBS can remain well in mesoporous channels when incubated with cells in culture medium.

As shown in Figure M, UCNPs@SiO₂@Au@PMO exhibits strong ultraviolet (UV) and blue emissions under 980 nm laser excitation. From a spectroscopic perspective (Figure N), upon 980 nm NIR illumination, photons are absorbed by Yb³⁺ ions and transfer energy to Tm³⁺ ions, resulting in strong blue emission at 475 nm (¹G₄ →³H₆). Other emissions at 365, 450, and 650 nm are attributed to the transitions ¹D₂ → ³H₆, ¹D₂ → ³F₄, and ³F₂ → ³F₄ of Tm³⁺ ions, respectively. These emission signals overlap well with the intense absorption of RBS in the 300–500 nm range (cyan dashed line in Figure M), enabling efficient energy transfer from UCNP to RBS salt. Additionally, Au nanoseeds exhibit strong absorption of 808 nm light (yellow dashed line in Figure M), which means that the UCNPs@SiO₂@Au@PMO-RBS can effectively absorb 808 nm light and transforms it to thermal heat.

2.2. Dissimilar Photon-Response of UCNPs@SiO₂@Au@PMO-RBS under 980 and 808 nm Light Excitation

Due to its unique dissimilar optical properties, the UCNPs@SiO₂@Au@PMO-RBS nanovehicle enables a “push–pull” response to two different NIR lights (i.e., 980 and 808 nm), where one increases particle speed as an accelerator while the other slows it as a brake. For example, upon illumination by 980 nm light, the UCNPs emit UV-blue light that activates RBS salt to release NO gas to propel the particle. In contrast, the 808 nm light absorbed by the Au seeds can generate a localized temperature gradient to slow the particle motion as a counteracting force.

First, the photothermal properties of the nanovehicle were investigated using 808 nm light as the excitation source. As shown in Figure A, the temperature of the UCNPs@SiO₂@Au@PMO-RBS solution can be rapidly increased by about 40 °C upon exposure to the 808 nm laser (1.5 W/cm², 15 min), significantly higher than that of pure water under the same irradiation condition (Figure A). To further calculate the photothermal conversion efficiency of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle, temperature variations were recorded under 808 nm laser irradiation at 1.5 W/cm² (Figure B). The fitted time constant was calculated as 592.6 s (Figure S16), corresponding to a photothermal conversion efficiency (η) of 24.1%. These results confirm that the UCNPs@SiO₂@Au@PMO-RBS nanovehicle exhibits good photothermal properties under 808 nm laser excitation. In contrast, 980 nm light irradiation only results in a much smaller temperature rise (Figure S17), indicating that heat generation is indeed primarily driven by 808 nm excitation.

2.

Photothermal effect and NO release of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle. (A) Temperature elevation curves of UCNPs@SiO₂@Au@PMO-RBS solution and water under different power densities of 808 nm laser irradiation (0.5, 1.0, 1.5 W/cm²) for 15 min. (B) Monitored temperature changes over 45 min with UCNPs@SiO₂@Au@PMO-RBS solution after 808 nm laser irradiation (1.5 W/cm²) for 15 min, following the laser being turned off. (C) Simulation of the thermal gradient distribution in the UCNPs@SiO₂@Au@PMO-RBS artificial nanovehicle in water under 808 nm light irradiation at 20 mW/cm². The corresponding distribution of the thermophoretic force is along the central line of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle. (D) The calibration curves of absorbance versus NO concentration. (E) NO release from the UCNPs@SiO₂@Au@PMO-RBS nanovehicle controlled by 980 nm light over time.

Subsequently, the heat distribution of UCNPs@SiO₂@Au@PMO-RBS was simulated, demonstrating an asymmetrical thermal distribution around the nanovehicle (Figure C). This is anticipated because the asymmetric distribution of Au seeds on one side of the UCNPs@SiO₂@Au@PMO-RBS can generate a rather localized temperature gradient via the thermophoresis process. When a nanomotor is subjected to a temperature gradient, the thermophoretic force (F _th) initiates its movement. As the particle accelerates, it experiences a counteracting Stokes drag force (F _d) from the surrounding fluid. When these two forces balance, the nanomotor moves at a constant thermophoretic velocity. , According to Stokes’ Law, the drag force (F _d) can be calculated as

$$F_{\mathrm{d}}=6\pi \eta rv$$ 2

where η is the dynamic viscosity of the fluid; r is the radius of the spherical object; and v is the velocity. Therefore, the thermophoretic force, F _th, can also be written as

$${F_{\mathrm{t}\mathrm{h}}=F}_{\mathrm{d}}=6\pi \eta rv$$ 3

For example, the velocity of the nanomotor can reach 0.60 μm/s at a power density of 20 mW/cm², corresponding to the thermophoretic force of 0.0012 pN (Figure S18).

Next, the NO release upon excitation with 980 nm light was investigated. According to the calibration curves of absorbance versus the NO concentration (Figure D), the amount of NO release can be obtained over time. As shown in Figure E, irradiation with 980 nm light results in a sharp increase in the NO concentration. However, once the 980 nm light is turned off, the generation and concentration of NO gradually decrease, further confirming that the NO release can be easily activated and controlled by 980 nm light.

2.3. “Accelerating-Braking” Motion Behavior of UCNPs@SiO₂@Au@PMO-RBS under Different Excitation Schemes

As mentioned above, the UCNPs@SiO₂@Au@PMO-RBS nanovehicle can release NO gas or localized temperature gradient upon excitation by 980 nm light or 808 nm light, which, in turn, can act as the driving force to propel the nanovehicle. Therefore, the motion behavior of UCNPs@SiO₂@Au@PMO-RBS under 980 or 808 nm was investigated, respectively. As shown in Figure A and Video S1, under low-power 980 nm laser excitation (6 mW/cm²), the nanovehicle moves at a relatively low velocity and only travels about 85.5 μm over 500 s (s). However, as the laser power increases, the motion becomes more directional and the travel distance significantly increases. For example, at a power density of 45 mW/cm², the total distance can already reach 478.1 μm over 500 s. A similar trend can also be observed under 808 nm light excitation, where the power significantly promotes the particle movement (Figure B and Video S2). The dependence of the nanomotor velocity on the NIR power density had been investigated (Figure S19A,B). It shows that the velocity of the nanomotor gradually increases as the power density of 808 and 980 nm NIR light increases. To clarify how the motor velocity changes in PBS media to demonstrate biological relevance, we have systematically compared the velocity of nanovehicles in both water and PBS media under 980 and 808 nm light irradiations. As shown in Figure S20, the propulsion velocity of the nanovehicle in PBS is slightly reduced in pure water under same illumination conditions. This can be attributed to the higher ionic strength and viscosity in PBS media, which may affect the propulsion movement. Nevertheless, the velocity in PBS still remains comparable, although slightly lower than in water, indicating that the propulsion behavior of the nanovehicles is still well retained in physiological environments. All these results demonstrate the potentials of the as-developed nanovehicles for biomedical applications.

3.

Acceleration and braking of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle. Time-lapse motion trajectories of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle under (A) 980 nm (6, 30, 45 mW/cm²) or (B) 808 nm (8, 12, 20 mW/cm²) light excitations. (C) Schematic illustration of light-controllable of NO release and thermal heating under 980 or 808 nm light excitations. (D) Time-lapse motion trajectories of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle under only 980 nm light excitation (blue line), 980 nm + 808 nm light excitation (red line), and a second round of 980 nm light excitation with cessation of 808 nm light (green line). The corresponding (E) velocity and (F) distance are obtained from trajectories in Figure D. The concentration of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle was 20 μg/mL.

Meanwhile, it is rather important to identify the propulsion mechanism of the nanovehicle upon release of NO under 980 nm light irradiation. For the bubble-recoil mechanism, propulsion mainly relies on momentum transfer during bubble detachment, which is largely insensitive to the ionic strength of the medium. However, in diffusiophoresis, the driving force arises from solute concentration gradients, which induce near-surface fluid flow (diffusio-osmosis) around the particle and causes it to move opposite to this interfacial flow. Therefore, solute gradients have little influence on bubble-recoil propulsion, but can significantly affect diffusiophoretic motion. For example, the “salt-suppression” effect is frequently observed in many types of nanomotors driven by self-diffusiophoresis or self-electrophoresis, such as α-Fe₂O₃, Pt, or multisegment Pt–Au–Ag_flex-Ni nanowire in H₂O₂ solution. − This is because the propulsion in these systems relies heavily on localized electric fields and the electrostatic electrical double layer (EDL). When external salt is added, the electrolytes severely compress the Debye screening length, thus dramatically reducing the propulsive driving force. , In light of these considerations, we have systematically compared the velocities of nanovehicles in solutions with varying NaCl concentrations. As shown in Figure S21, increasing the NaCl concentration adequately reduces velocity. This is because NO-driven propulsion depends on the interaction potential between NO molecules and the particle surface within a thin interfacial exclusion layer (1 ∼ 10 nm). If high concentration of NaCl is added to the medium, the accumulated Na⁺ and Cl^– ions near the surface can perturb the interfacial water structure due to their strong hydration shells, which weakens the NO-surface interaction potential and reduces the velocity. Besides, the addition of NaCl salt, particularly at high concentration, increases the viscosity and thermal conductivity of the liquid media, which may further reduce the propulsive velocity at a given force. Taken together, it is concluded that the propulsion via NO release under 980 nm light irradiation is dominated by diffusiophoresis rather than bubble recoil.

Moreover, due to the specific arrangement of RBS salt and Au seeds (Figure C), it is proposed that the generated driving forces from RBS and Au may also counteract each other, thus achieving a precise control of speed and direction. In order to verify this concept, the time-dependent movement trajectories of UCNPs@SiO₂@Au@PMO-RBS nanovehicles were further studied under different excitation schemes, i.e., excitation by only 980 nm or simultaneous excitation using both 980 and 808 nm NIR lights (980 nm + 808 nm). As shown in Figure D and Video S3, under 980 nm laser excitation alone (blue), the nanovehicle exhibits directional motion with a speed ranging from 0.4 to 0.8 μm/s for the first 256 s (Figure E). However, upon switching on the 808 nm laser (red), its propulsion is significantly interrupted and velocity is reduced to 0 ∼ 0.2 μm/s (Figure E). Then, once the 808 nm light is switched off at the time point of 512 s (green), the nanovehicle resumes its motion with its velocity increased to 0.4 ∼ 0.6 μm/s. Additionally, the corresponding cumulative displacement over time is also shown in Figure F. During the initial period from 0 to 256 s when only 980 nm illumination is applied, the cumulative distance of the nanovehicle steadily increases. However, from 256 to 512 s when an additional 808 nm light is introduced, the cumulative distance remains nearly unchanged, indicating a pronounced braking effect. Finally, from 512 to 600 s, with the removal of 808 nm light excitation (i.e., only 980 nm light is used), the cumulative distance is observed to increase over time again. To verify active motion at the nanoscale, we have provided Mean Squared Displacement (MSD) curves based on motion trajectories of 10 individual particles under the three conditions (Figure S22 A,B): 980 nm light on (“Go” state), both 980 and 808 nm light on (“Stop” state) and no light on (“Dark” state). As shown in Figure S22C, the MSD curves exhibit a roughly parabolic shape for both “Go” and “Stop” states, which clearly indicate directional motions. , Additionally, the MSD values under “Stop” state are much lower than in the “Go” state, suggesting a substantial decrease in propulsion velocity. Therefore, it can be concluded that the addition of 808 nm irradiation effectively suppresses the propulsion speed as an optical brake. As a comparison, the MSD curve in the absence of light irradiation presents a linear correlation with time intervals (Figure S22D), indicating the movement with no light irradiation is Brownian-type.

Notably, such nonlinear and parabolic-shaped MSD curves in Figure S22C at time scales exceeding the orientation diffusion time have also been observed in other similar nanomotor systems. − Specifically, Cao et al. and Wang et al. once attributed this feature in the Au-mediated nanomotors to the highly localized heating and formation of asymmetric thermal gradients. , In our study, we hypothesized that the parabolic shape of the MSD curve could be due to the intrinsic structural and material asymmetry of the nanovehicle particle: since the Au nanoparticles are coated asymmetrically only on one side of the SiO₂ shell, this would generate a nonuniform temperature gradient in the surrounding solution rather than a uniform thermal field. Such a localized temperature field induces strong thermophoretic forces that drive directional particle motion. Moreover, the propulsion generated by localized heating can be sufficiently strong to compete with or effectively overcome the intrinsic rotational diffusion process. This can also be verified by the motion trajectories of nanomotors in Figure S22A, which also rules out the possibility of convective flow as individual particle trajectories are noncorrelated in direction.

Taken together, all these results demonstrate a high-spatiotemporal control over nanovehicle motions have been achieved by leveraging photothermal and photochemical propulsion mechanisms. Specifically, by managing the “on–off” switching of 980 and 808 nm NIR light sources to induce acceleration or braking, the velocity and direction of the nanovehicle can be finely tuned, representing an advanced level of motion regulation with broad implications.

2.4. Controlled Cellular Uptake and Programmed Gas/Photothermal Therapy

On the basis of such precise motion control of the nanovehicle, its potential application for cell–particle interactions and programmable phototherapy was further explored. Previous studies have reported that the motion of nanoparticles can affect nanoparticle–cell interactions, , and thus can be used to regulate cellular endocytosis. In another study, Zhang et al. once used dissipative particle dynamics simulations to investigate the influence of the nanoparticle speed on the integrity of cell membranes. According to their simulation, the rotation of nanoparticles could provide a strong driving force for membrane rupture. Meanwhile, anisotropic nanoparticles were also found to exhibit higher membrane rupture efficiency than the symmetric ones. HeLa cells were first cocultured with UCNPs@SiO₂@Au@PMO-RBS for 24 h and subsequently analyzed (Figure B). The blue signal indicates the presence of cell nucleus while the cyan signal corresponds to upconversion emissions (400 ∼ 500 nm) from nanovehicle. The results show that under 980 nm light excitation, a stronger intracellular upconversion signal is detected compared with blank group, suggesting enhanced cellular uptake. However, when both 980 and 808 nm NIR lights are applied simultaneously, the upconversion signal is significantly reduced, indicating cellular uptake is prohibited. This difference in uptake is attributed to the motion state of the nanovehicle. Under 980 nm light excitation, NO gas is generated and propels the nanovehicle, which increases its interaction with the cell membrane and thereby facilitates endocytosis. In contrast, when both 980 and 808 nm lights are applied, the slowed velocity due to braking effect from 808 nm light illumination weakens nanoparticle–cell interactions, thus leading to decreased cellular uptake.

4.

Light-controlled endocytosis and in vitro study. (A) Schematics of light-controlled endocytosis of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle. (B) Luminescence images of HeLa cells incubated with the UCNPs@SiO₂@Au@PMO-RBS nanovehicle (blank), and excited by either a 980 nm light (980 nm, 45 mW/cm²) or simultaneous dual-light irradiation (980 + 808 nm, 980 nm, 45 mW/cm²; 808 nm 20 mW/cm²). (C) Cell viability of HeLa cells incubated with the UCNPs@SiO₂@Au@PMO-RBS nanovehicle at different concentrations. (D) Viability of HeLa cells incubated with the UCNPs@SiO₂@Au@PMO-RBS nanovehicle under different NIR light combinations. (E) Luminescence imaging of ROS production. The experiment is composed of six groups: (1) control group, receiving an injection of PBS (PBS); (2) injection of the drug followed by irradiation with 980 nm light (N); (3) injection of the drug followed by irradiation with 808 nm light (P); (4) injection of the drug with simultaneous irradiation of 980 and 808 nm light (N + P); (5) injection of the drug with sequential irradiation, first with 808 nm light and then with 980 nm light (P → N); and (6) injection of the drug with sequential irradiation, first with 980 nm light and then with 808 nm light (N → P). (F) Luminescence imaging of intracellular NO production at different times after irradiation by 980 nm light (45 mW/cm²). The concentration of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle was 200 μg/mL.

Additionally, nitric oxide (NO) has long been recognized as a key signaling molecule involved in regulating diverse physiological processes, such as neurotransmission, vasodilation, and gene expression. − While NO plays essential roles in normal physiology, its overproduction has been associated with cellular dysfunction and damage. Considering its ability to release NO gas, the application of the nanovehicle in programmable phototherapy against tumor cells was further investigated. Before performing in vitro and in vivo phototherapy, it is important to evaluate the cytotoxicity of UCNPs@SiO₂@Au@PMO-RBS nanovehicles. HeLa cells were incubated with varying concentrations of nanovehicle (0, 25, 50, 100, and 200 μg/mL) for 24 h, and cellular viability was assessed using the standard CCK-8 assay. As shown in Figure C, the presence of UCNPs@SiO₂@Au@PMO-RBS has neglectable effect on cell viability, suggesting its good biocompatibility. Meanwhile, the effect of the laser irradiation on the cell viability was also investigated. As shown in Figure S23, the viability of HeLa cells can still maintain more than 90% under 980 and 808 nm NIR lights with different power densities, indicating that the used laser power has neglectable effect on cell viability. Then, to evaluate the treatment efficiency of the designed nanovehicle, HeLa cells were incubated with UCNPs@SiO₂@Au@PMO-RBS and then exposed to 980 and 808 nm light in different combinations (Figure D). Upon 980 nm light excitation (N group), cell viability was reduced to about 80%. This is likely due to NO gas production, which disrupts cellular respiration and generates cytotoxic free radicals. When the excitation is switched to 808 nm light (P group), cell viability is observed to decrease to approximately 60% due to the photothermal effect. A similar influence of light irradiation on cell viability can be found in other groups, including 980 + 808 nm (N + P group) and 808→980 nm (P → N group). However, the viability is reduced most for HeLa cells treated with 980 nm light, followed by 808 nm light (N → P group), where the cell viability only remains about 20%. Thus, it can be concluded that stepwise light irradiation using the NO/photothermal sequence causes the most significant damage to HeLa cells. This is because under 980 nm light excitation, the prerelease of NO can directly induce mitochondrial/DNA damage and block DNA synthesis and repair, thereby impairing the autophagic self-repair ability of tumor cells. Meanwhile, the NO can diffuse to surrounding cells, which can further mitigate limitations of uneven heat distribution in PTT upon switching to 808 nm light, and further disrupt cellular self-repair. −

To effectively visualize intracellular NO and ROS production, HeLa cells treated with UCNPs@SiO₂@Au@PMO-RBS were coincubated with the DCFH-DA and DAF-FM-DA fluorescent probes. Figure E compares intracellular ROS signals under different excitation combinations. Among them, the N → P group displays the highest ROS signal, indicating a more effective ROS-generating mechanism under this specific sequence. For example, according to the fluorescence intensity analysis, the ROS signal in the N → P group is 3.3 times higher than that in the P → N group and 3.8 times higher than that in the N + P group. The electron paramagnetic resonance (EPR) measurements were used to determine the type of generated ROS with DMPO (5,5-dimethyl-1-pyrroline-N-oxide) as the spin-trapping agent. The obtained EPR spectrum clearly exhibits characteristic signals of superoxide radical (^•O₂ ^–) (Figure S24). Furthermore, Figure F shows that NO fluorescence still remains weak during the initial 15 min of 980 nm light excitation and only reaches maximum after 60 min of continuous exposure. This delayed accumulation also suggests that simultaneous activation of NO release and photothermal heating may prevent NO from effectively diffusing and disrupting cellular respiration. Therefore, these findings furthermore highlight the necessity of a stepwise activation strategy, allowing sufficient time for NO to exert its biological effects before heat generation.

2.5. In Vivo Gas/Photothermal Synergistic Therapy

The therapeutic efficacy of UCNPs@SiO₂@Au@PMO-RBS for in vivo phototherapy was systematically assessed using a small animal model. To establish the tumor-bearing mouse model, HeLa cells were subcutaneously injected into the right axilla of BALB/c mice. The therapy experiments were initiated once the tumor volume reached ≈100 mm³. Subsequently, UCNPs@SiO₂@Au@PMO-RBS was delivered into the mice via intratumoral injection. To prevent excessive heating, the mice were exposed to laser irradiation for 5 min, with a 3 min interval between each exposure. The infrared thermal images presented in Figure S25, which illustrate that the surface temperature remains below 37.2 °C for a 5 min duration under irradiation by 980 nm (PBS + 980) and 808 nm (PBS + 808) light. It is indicated that the thermal effect induced by laser irradiation only is negligible. However, after injection with UCNPs@SiO₂@Au@PMO-RBS, significant heat generation is observed at the tumor site under irradiation with 808 nm light or with both 980 and 808 nm light, indicating that UCNPs@SiO₂@Au@PMO-RBS can effectively elevate the tumor temperature for PTT upon illumination by NIR light.

For the therapeutic treatment, HeLa tumor-bearing mice were randomly assigned to six experimental groups: (1) control group, receiving an injection of PBS (PBS); (2) injection of the drug followed by irradiation with 980 nm light (N); (3) injection of the drug followed by irradiation with 808 nm light (P); (4) injection of the drug with simultaneous irradiation of 980 and 808 nm light (N + P); (5) injection of the drug with sequential irradiation, first with 808 nm light and then with 980 nm light (P → N); and (6) injection of the drug with sequential irradiation, first with 980 nm light and then with 808 nm light (N → P). During the treatment period, tumor size and body weight of the mice were recorded at two-day intervals following the initial treatment. As shown in Figures A and S26A, a continuous increase in tumor size is observed in mice from control group 1 (PBS), and group 2 (N). In contrast, the group 3 (P), group 4 (N + P), or group 5 (P → N) exhibit moderate therapeutic effect. Notably, the most significant tumor inhibition is observed in group 6 (N → P) with an inhibitory rate about 80%. The body weight of mice in all six experimental groups show a consistent increase throughout the 14 day treatment period, as shown in Figures B and S26B. This increase indicates that the treatment has no apparent adverse effects on the overall health condition of the mice, suggesting that the UCNPs@SiO₂@Au@PMO-RBS nanovehicle does not induce significant toxicity. Furthermore, the representative images of excised tumors (Figure C) provide direct visual evidence that the smallest tumor size is achieved in group 6 (N → P).

5.

In vivo study. Relative tumor volume growth (A) and relative mouse body weight (B) in different groups after various treatments. Photographs of isolated tumors (C), H&E stained images (D) and TUNEL analysis (E) after treatment for 14 days. The concentration of the UCNPs@SiO₂@Au@PMO-RBS nanovehicle was 200 μg/mL.

To investigate the primary mechanisms of phototherapy treatment, histological analyses were performed on tumors and major organs across different experimental groups after the 14 day post-treatment period (Figure D). The results indicate that the N → P group has the most pronounced levels of apoptosis and necrosis. As shown in Figure E, a similar trend is observed in the TUNEL (Terminal deoxynucleotidyl transferase dUTP Nick-End Labeling) assay, which demonstrates the most significant suppression of tumor proliferation in this group. Furthermore, hematoxylin and eosin (H&E) staining of major organs (heart, liver, spleen, lung, and kidney) from mice in different treatment groups (Figure S27) reveals no apparent organ damage or pathological abnormalities. These results suggest that the developed UCNPs@SiO₂@Au@PMO-RBS nanovehicle exhibits no immediate systemic toxicity. Blood hematology analysis was conducted on mice following various treatments (Figure S28), and no acute changes were observed among the different groups.

3. Conclusion

In this study, we reported an artificial nanovehicle, UCNPs@SiO₂@Au@PMO-RBS, with counterbalancing-driving mechanisms for precise control over its speed and direction. Upon excitation by 980 nm light, the UCNP emits blue light to activate RBS salt to release NO gas and propel the particle at a velocity of 0.8 μm/s. When the light is switched to 808 nm light, Au nanoseeds generate a localized thermal gradient in the opposite direction, functioning as an optical brake. Due to the independence of NO release and thermal heating, the speed and direction of the nanovehicle can be precisely manipulated under selective light excitation. As a proof of concept, such motion control was used for regulation of cellular endocytosis and programmable phototherapy. The results show that the intracellular uptake can be noninvasively and remotely controlled by alternating 980 and 808 nm NIR light excitations. Besides, thanks to synergistic effect of NO-induced cellular dysfunction and photothermal conversion of Au, the nanovehicle further demonstrates promising applications for programmable phototherapy with a tumor inhibition rate about 80%. It is believed that the dual-light-responsive nanovehicle demonstrates great potentials for controllable endocytosis and programmable therapy, offering a promising platform for precise clinical applications..

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.6c00299.

• Video S1: The movement of nanovehicle under illumination of 980 nm light (AVI)

• Video S2: The movement of nanovehicle under illumination of 808 nm light (AVI)

• Video S3: The movement of nanovehicle under illumination of 980 nm and 808 nm light (AVI)

• Experimental details, characterization methods, investigation of propulsive mechanisms, as well in vitro and in vivo studies of UCNPs@SiO₂@Au@PMO-RBS nanovehicles (PDF)

∥.

J.W., J.L., and M.T. contributed equally to this work.

The authors declare no competing financial interest.