Ying Wangab,
Zhen Lib,
Alexander A. Soloveva,
Gaoshan Huang*a and
Yongfeng Meia
aDepartment of Materials Science, Fudan University, Shanghai 200433, People's Republic of China. E-mail: gshuang@fudan.edu.cn
bDepartment of Physics and Mathematics, Shanghai University of Electric Power, Shanghai 201300, People's Republic of China
First published on 17th September 2019
In this work, UV light-controlled two-dimensional (2D) TiO2 plate micromotors are demonstrated for the first time. The 2D TiO2 micromotors are produced by the well-known anodic oxidation method in combination with a cracking and separation process. When the motor is placed in H2O2 aqueous solution under UV irradiation, oxygen bubbles are generated in the holes of the TiO2 membrane. The 2D micromotor thus moves upon O2 bubbles under its own weight. In contrast to bubble-propelled micromotors, which require an addition of surfactants to chemical fuels, the 2D micromotor is capable of moving in aqueous H2O2 solution without surfactants. Moreover, speed of the 2D TiO2 micromotor can be controlled by the intensity of the UV light. Such surfactant-free micromotors and their facile fabrication hold considerable promise for diverse practical applications in the biomedical and energy fields, for example, and in new materials.
Many different kinds of photo-activated micro-/nano-motors have been developed so far. The geometries of the motors include small particles (i.e., 0D structures),22 rods (i.e., 1D structures),23 tubes,9,11,24 and capsules (i.e., 3D structures).25,26 For example, Dong and coworkers demonstrated the efficient propulsion of UV light-driven TiO2–Au Janus micromotors in water.27 Zhou et al.28 described a visible light-driven microparticle structured Cu2O–Au micromotor that displayed autonomous motion in low concentration H2O2 solution. Mou and co-workers demonstrated tube-structured light-controlled micromotors.29 Wong et al.30 have developed light-activated AgI–Pt nanorod micromotors and Ag–Pt micropump systems. Wu and coworkers described polymer-based microtubular engines activated by near-infrared laser.19 Most of the light-driven micro-/nano-motors reported to date are of 0D, 1D, and 3D micro-/nano-structures, and compared with these structures, the plate-like porous 2D micromotors have high loading capacity.31 Recently, crystalized porous TiO2 membrane-based motor was introduced as the first demonstration of 2D porous nanomotor, where ∼30 particles were loaded.31 In general, amorphous TiO2 is inactive under UV irradiation, because the photo-generated electrons and holes recombine with each other efficiently.32 However, when Am-TiO2 was sensitized by H2O2, it possess a superior photocatalytic activity compared with the crystalline TiO2, and the surface peroxide complexes can effectively enhance separation ability for the photo-generated charge pair.32 Thus, light-driven micromotor fabricated from amorphous TiO2 may possess enhanced performance while the related research is still scarce.
In addition, the presence of surfactant in the environment is generally necessary for the movement of most bubble-propelled micro-/nano-motors, as it lowers interfacial tension required for detachment of the bubble from the micro-/nano-motors.33 However, it is considered that an addition of surfactant is not biocompatible and can be harmful for environment.4 Therefore, micro-/nano-motors that can efficiently move in fluids without surfactant are highly desired.
In this work, we design amorphous TiO2 2D micromotors (Am-TiO2 2D micromotors) by separating anodic TiO2 membranes into bunches of TiO2 nanotubes. We demonstrate that these Am-2D-TiO2 micromotors can be driven by UV light via a so-called bubble-supported gravitational propelling. We notice that the motion mode of Am-TiO2 2D micromotor is entirely different in comparison to previously observed bubble-propelled mechanism. For the bubble-propelled motors, the moving directions of the motor and the bubble are opposite.8,9,34,35 However, the Am-2D-TiO2 micromotors moved upon the generated bubbles. During the movement, the photo-generated O2 molecules firstly accumulate in the holes at the bottom of the motor and grow into bubbles with different radii. The pushing force from the bubbles makes the 2D micromotor float on the bubbles with an angle between the 2D micromotor and horizontal plane, due to different radii of the bubbles. Then, the gravity of the 2D micromotor causes the motor move along the direction from larger bubble to small bubble. Our work demonstrates that when the motor is exposed to UV light, the bubble-supported gravitational propulsion is immediately triggered which efficiently drives the micromotors in H2O2 solution without any surfactants. Moreover, the motor can be intentionally switched on or off by the light source, and the velocity of Am-TiO2 2D micromotor can be adjusted on demand by altering the light intensity.
Fig. 1 (a) SEM image of the sample. (b) EDX spectrum and XRD pattern (inset) of the as-prepared sample. |
The efficient photocatalytic H2O2 decomposition over the amorphous TiO2 under UV irradiation plays a crucial role in the light-controlled propulsion of the Am-TiO2 2D micromotors. Usually, for TiO2, only crystalline phases such as anatase, rutile, and anatase/rutile mixed phases are photocatalytically active under UV irradiation.29,36 Amorphous TiO2 has high density of defects,32 and the facilitated recombination of photogenerated electrons and holes makes amorphous TiO2 nearly inactive under UV irradiation.32 However, when amorphous TiO2 is sensitized by H2O2, the amorphous TiO2 exhibits an efficient photocatalytic activity even higher than that of crystalline TiO2.37–39 Amorphous TiO2 decomposes the H2O2 fuel in the presence of UV light according to equations:29,40
TiO2 + hν → h+ + e− | (1) |
H2O2 + 2h+ → O2 + 2H+ | (2) |
H2O2 + 2e− + 2H+ → O2 + 2H2O | (3) |
Video 1, ESI† and Fig. 2 show the photoactive motion of the Am-TiO2 2D micromotor in 10 wt% H2O2 without any surfactant, and the power density of UV light is 0.5 W cm−2. Once the UV light is on, the Am-TiO2 2D micromotor moves upon O2 bubbles with diameter of ∼50 μm (Fig. 2b–d), which comes out from the bottom side of the 2D micromotor. Although the bubble generation rate is 2 s−1, a remarkable motion speed of 53 μm s−1 is observed for the 2D micromotor. When the UV light is off, the 2D micromotor stops immediately (Video 1, ESI† and Fig. 2c). When the UV light is turned on again (Fig. 2d), the 2D micromotor can be reactivated within 0.5 s, indicating the fast response ability of the 2D micromotor upon the UV irradiation. Compared with bubble-propelled catalytic micromotors, the present 2D micromotor demonstrates a different moving mode. From Video 1, ESI,† one can see that the bubbles come out from the bottom, while the Am-TiO2 2D micromotor moved laterally. The observed phenomenon implies a new moving mechanism for the 2D micromotors, which will be discussed in detail later.
Fig. 2 Time-lapse images of the light-controlled motion of the Am-TiO2 2D micromotor in 10 wt% H2O2 at: (a) 3 s, (b) 6 s, (c) 9 s, and (d) 12 s. The orange arrows indicate the Am-TiO2 2D micromotor. The green lines in (a), (b), (c), and (d) correspond to motion trajectories at time slots of 0–3, 3–6, 6–9, and 9–12 s, respectively. The micromotors are irradiated by the UV light with a power density of 0.5 W cm−2, and the UV light is on during 3–6 s and 9–12 s (Video 1, ESI†). Scale bars: 200 μm. |
To explore the factors which can influence the motion of Am-TiO2 2D micromotor, we studied the motion speed of the micromotor under different light intensities and H2O2 concentrations. Video 2, ESI† and Fig. 3a reveal that when power density of the UV light increases from 0.2 to 0.5 W cm−2, the speed of the 2D micromotors increases from 23 to 52 μm s−1. A linear relationship between the motor speed and the light power density is observed (Fig. 3a). However, there is no motion observed for the 2D micromotors, when UV light power density is less than 0.1 W cm−2, and bubbles ejection is sparsely observed. This indicates that a lower limit of the UV light power density should exist, and the motor begins to move when the light power density is larger than this value. Video 3, ESI† and Fig. 3b show the moving velocities of Am-TiO2 2D micromotor in H2O2 solutions with different concentrations. The speed of the micromotors is ∼24 μm s−1 in 1 wt% H2O2 and it increases to 49 μm s−1 in 8 wt% H2O2. However, as the concentration increased from 8 wt% to 10 wt%, the speed increase slowly. One can see that the velocity is not linearly proportional to the H2O2 concentration in the range of 1–10 wt%, suggesting that the energy conversion efficiency of the catalytic motion is rate-limiting.28
In order to better understand the motion mechanism of the Am-TiO2 2D micromotor under UV illumination, we analyzed the motion in details and six bubbles are specifically tracked (see Video 4, ESI†). Fig. 4a–f show the release of six O2 bubbles produced by the micromotor. Fig. 4a shows three bubbles with different sizes, which are indicated as 1, 2, and 3 respectively, and the radii of the bubbles are R1 > R2 > R3. In the present case, the different bubble size is considered to be due to the growth of bubble in the holes with different sizes (Fig. 1a), which can be used to store gas and produce bubbles. Firstly, the 2D micromotor moves upon the bubbles, along the direction from bubble 1 to bubble 2 (Fig. 4b–c). Then, the 2D micromotor moves from bubble 2 to bubble 3, leaving bubbles 1 and 2 behind (Fig. 4d). Here, bubble 1 merges into another bubble, and its initial position is marked with a dashed circle. At the same time, bubble 4 (R4 < R3) forms at the bottom of the 2D micromotor (Fig. 4d). When the 2D micromotor moved from bubble 3 to bubble 4, bubble 5 and bubble 6 appear (Fig. 4e and f). The formation of the three bubbles (bubbles 4–6) with different sizes (R4 > R5 > R6) at the same positions as those of bubbles 1–3, suggesting another moving cycle, and therefore the motion continues.
The UV-induced motion of the Am-TiO2 2D micromotor in H2O2 fuel is schematically illustrated in Fig. 4g and it can be divided into four stages. Firstly, without UV irradiation, no bubble or movement is generated for the 2D micromotor (Stage 1, t = t0). For the sake of simplicity, we assume that there are two holes with different sizes at the bottom side of the 2D micromotor, which are signed as A and B in Fig. 4g. When a UV light is on, the photogenerated O2 molecules can accumulated in these holes. When the holes are full of O2 molecules, two bubbles with different radii (R1 > R2) are formed, as shown in Stage 2 of Fig. 4g. In Stage 3, the pushing force N from the bubbles make the 2D micromotor float on them. Because of different radii of the bubbles, a tilt angle θ between the surface of 2D micromotor and horizontal plane appears, and gravity G can be divided into two components G1 and G2. As illustrated in Fig. 4g, the G1 and G2 can be expressed as:
G1 = Gsin(θ) = Mgsin(θ), | (4) |
G2 = Gcos(θ) = Mgcos(θ) = N, | (5) |
In Stage 3, the gravity component G1 makes the 2D micromotor moving along the G1 direction. Then, the system evolves to Stage 4, and the movement in one step with a distance of L is accomplished. One can see that for movement in one step, the micromotor moves from big bubbles to small bubbles (Fig. 4g), and the process cycles to achieve continuous movement. Based on above eqn (4) and (5), we conclude that the average velocity for 2D micromotor can be written as
(6) |
The surfactant plays a major role in reducing the surface tension of the fuel and generating bubbles at high rate.41,42 Therefore, the velocity of bubble-propelled micro-/nano-motors in previous studies is remarkably higher if surfactants are added.33 However, the present Am-TiO2 2D micromotor demonstrated different motion behaviors, and the surfactant could not promote the locomotion of such micromotors. On the contrary, we found that the surfactant could hinder the movement of the Am-TiO2 2D micromotor. In our experiment, we studied the influence of surfactant on the motion behavior by observing the locomotions of Am-TiO2 2D micromotors in solutions containing sodium dodecyl sulfate (SDS) with different concentrations (0, 0.3, 1, 2, and 5 wt%). Video 5, ESI† and Fig. 5a show that without surfactant, the bubbles with different sizes start to grow at the bottom side of the 2D micromotor once the UV irradiation is on. The 2D micromotor then effectively moves upon the generated bubbles. On the contrary, low SDS concentration (i.e., 0.3 and 1 wt%) led to slightly decreased motion speed. After 2 wt% or more SDS being added, both the number of the bubbles and the bubble production rate are increased. However, the 2D micromotor can hardly move, as shown in Fig. 5b. With further increase of the SDS concentration to 5 wt%, the 2D micromotor cannot move, although more bubbles come out from the bottom (Fig. 5c). It is therefore concluded that for the 2D micromotor the surfactant cannot promote, but make against the efficiency of locomotion. This phenomenon is further discussed below.
On the basis of above bubble-supported gravitational propelling mechanism for the 2D micromotor (Fig. 4g), we are able to specifically explain the influence of surfactant on the locomotion. Generally, the motion mode of traditional bubble-propelled motors is related to jet engine mechanism, which propels micromotors in opposite direction with respect to generated bubbles.43–47 The surfactant increases the bubble generation rate, and the micromotor speed increases correspondingly.41,42 In contrast, the Am-TiO2 2D micromotor moves upon the bubbles. In comparison with bubble-propelled motors, the Am-TiO2 2D micromotor needs larger bubbles to support the 2D micromotor, and bubbles with different sizes cause the motion of the motor (eqn (6)). Without surfactant, the O2 molecules can accumulated in the holes with different sizes, then form bubbles with different radii. It is important to notice that generation of bubbles with different radii is necessary for the 2D micromotor, because a tilt angle θ can thus be created, and component of the gravity (i.e., G1) leads to the movement of micromotor along this direction (Fig. 4g). However, upon an addition of surfactant, the O2 molecules cannot accumulate and large bubble are not generated at the bottom of the 2D micromotor due to reduced surface tension.41,42 As a result, although more bubbles come out from the bottom of the micromotor, the 2D micromotor does not move effectively (Fig. 5b and c).
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ra06426e |
This journal is © The Royal Society of Chemistry 2019 |