1. Introduction

Graphene is one of the most important materials in the past two decades due to its interesting properties such as high electron mobility [1], excellent mechanical strength [2], stress-induced pseudomagnetic field [3–5], superconductivity in twisted system [6] and so forth. When a light beam irradiates on graphene-based materials, both the direct momentum transfer and the photon-electron-atom energy transfer happen, which provides multiple physical principles to achieve light-driven manipulation of objects, namely the transformation from light to work.

One light manipulation strategy is based on the photothermal deformation of graphene materials due to thermal mismatch inside the devices upon light irradiation. For instance, light-assisted hand and robots based on graphene oxide artificial muscles were reported [7]. Besides, an artificial phototropism device for omnidirectional tracking and harvesting of light was presented by using the reduced graphene oxide and other photothermal responsive materials [8].

Different form the abovementioned deformation approach, light-driven direct translational motion of graphene-based materials is also investigated in recent years, in which a typical case is the light propulsion. For example, our group continuously studied the direct laser propulsion of graphene sponge (a kind of 3D-crossed graphene structure) in rarefied gas with variable vacuum degree, where the measured propulsion force is approximately three orders larger than the radiation pressure [9], the dynamic mechanism is more likely originated from the laser-induced nonequilibrium of interface molecular momentum transfer [10,11], and a fast propulsion is also realized by a pulsed laser [12]. Zhou et al. presented a near-infrared light-steered graphene aerogel micromotor actuated by asymmetric surface tension of water and achieved precise navigation for active transport and microassembly [13]. Strek et al. shown direct light-induced propulsion of vessels filled with a suspension of graphene particles and methanol, and they discussed the mechanism of propulsion effect in terms of the explosive hydrogen-oxygen reaction [14].

Apart from light-driven deformation [15] and propulsion, optical pulling is another light manipulation method that has been developed rapidly in the past ten years [16,17]. Several optical pulling mechanisms were proposed and verified in liquid mediums, such as opto-thermoelectric pulling [18], opto-thermophoretic pulling [19]. However, light-driven pulling of graphene in liquid medium is rarely reported. Optical pulling and collection of graphene in liquid has many potential applications. For example, this will provide an optical printing or optical assembly method to fabricate graphene optical devices and structures. Besides, this can be used in micronanofluidic devices to achieve optical sensing based on graphene materials [20–22].

In this work, massive laser pulling of GN in pure water is presented. The adopted light source is a common unfocused gauss laser beam. This pulling phenomenon holds for multiple laser wavelengths including 405 nm, 488 nm, 532 nm and 650 nm. The velocity information of GN is quantitatively analyzed. This work provides an effective massive light manipulation scheme for graphene material, which is valuable for nanofluidic and optomechanical devices.

2. Experimental setup and method

2.1 Preparation of the graphene aqueous solution

Multilayer graphene was mechanical exfoliated form a bulk highly oriented pyrolytic graphite (XFnano) by using a sticky tape (3 M) [23]. Then, the exfoliated graphene was transfered into the deionized water in a transparent water tank (15 cm × 11 cm × 13 cm) by using a tweezer. In the end, ultrasonic treatment was used to obtain a uniformly dispersed graphene aqueous solution, where the ultrasonic time is 15 minutes.

2.2 Size characterization of the GN

After the ultrasonic treatment, 15 µL graphene aqueous solution was dropped onto the surface of a clean glass slide and then dried in room temperature. An atomic force microscope (AFM, Hitachi, AFM5000II) was used to measure the width and height of the GN on the glass slide. The measurement areas were randomly selected.

2.3 Experimental method of laser pulling, motion monitoring, and motion analysis

Four laser light sources with different wavelengths including 405 nm (OXlasers), 488 nm (Coherent Sapphire), 532 nm (OXlasers) and 650 nm (OXlasers) were used to irradiate the graphene aqueous solution through the transparent tank wall with an incident angle of 90° as shown in Fig. 1. The output laser powers used in experiments were measured by a light power meter (THORLABS PM100A). All the laser lights are unfocused gauss beams with a plane wavefront. The real-time moving images of GN upon laser irradiation were enlarged by a microscope and then monitored by a CCD device connected with a computer (Fig. 1). This motion monitoring system was used to directly observe the laser pulling phenomena and also record the flow field in graphene aqueous solution. The frequently-used mathematical software MATLAB supplies a particle image velocimetry tool (PIVlab) [24], by which the instantaneous velocity information of GN upon laser irradiation is calculated based the real-time moving images recorded by the CCD device.

 

Fig. 1. Schematic diagram of the experimental setup.

Download Full Size | PDF

3. Results

3.1 Size characterization of the GN

According to the preparation method of the graphene aqueous solution stated in the section 2.1, the chemical components of the solution are very simple just including pure water and graphene. The dimension information of graphene is important in a laser irradiation process. As shown in Fig. 2(a) and (b), GN with irregular shapes are formed in water after the ultrasonic treatment, which possess random sizes in width ranging from tens of nanometers to several micrometers. The uniform image contrast in Fig. 2(a) and (b) indicates well layered structures of graphene. In order to obtain the detailed height information of GN, the surface profiles at the position 1, 2 and 3 (marked with white lines in Fig. 2(a) and (b)) were measured by AFM and shown in Fig. 2(c-e). According to the surface profiles, the thickness of GN at the position 1, 2 and 3 is approximately ∼ 5 nm. Because a single layer graphene is ∼ 0.33 nm in thickness [25], these nanosheets contains more than ten layers of graphene at least, which are capable to partially absorb the incident photon energy at the visible light band used in this work.

 

Fig. 2. AFM characterization of the GN. (a, b) Surface morphology of randomly-selected GN on a glass slide. (c-e) Cross-section analysis of GN at three different positions shown in a and b marked with white lines.

Download Full Size | PDF

3.2 Experimental confirmation of the massive laser pulling phenomenon

As shown in Fig. 3(a), a laser beam with wavelength of 488 nm and power of 80 mW was used to irradiate the graphene aqueous solution and the propagation direction of light is from left to right. The motion video (Visualization 1) of GN were recorded by a microscope and a CCD device, in which the collective movement of the GN from right to left is observed. In order to get more accurate motion information, the real-time velocity field of GN upon laser irradiation is calculated with the PIVlab software as shown in Fig. 3(b), where the green arrows represent the instantaneous velocity vectors of GN in the flow field. The velocity field in Fig. 3(b) undoubtedly proves that the GN have a remarkable velocity component toward the laser light source. In other words, massive laser pulling of GN in water is realized. Furthermore, this experiment was repeated by reversing the laser propagation direction (Fig. 3(c)). In this case, as shown in Visualization 2 and Fig. 3(d), the GN still move towards the laser light source. Therefore, the massive laser pulling phenomenon is confirmed.

 

Fig. 3. Experimental confirmation of the massive laser pulling phenomenon. (a) Laser pulling image of the GN upon 488 nm laser irradiation, where the laser propagation direction is from left to right. (b) Instantaneous velocity field of the GN in the dotted box area in a. (c) Laser pulling image of the GN, where the laser propagation direction is from right to left. (d) Instantaneous velocity field of the GN in the dotted box area in c. The scale bars in a and c are 0.4 cm.

Download Full Size | PDF

3.3 Massive laser pulling with different wavelength

In order to further verify the universality of the laser pulling phenomenon on graphene nanosheets, different laser parameters should be tested such as light wavelength and laser power. As shown in Fig. 4 (a), (c) and (e), unfocused laser beams with wavelength of 405 nm, 532 nm and 650 nm are used to irradiate the graphene aqueous solution, respectively, in which the incident laser power is 150 mW, 80 mW and 60 mW, respectively. The corresponding motion videos of GN upon laser irradiation are shown in Visualization 3, Visualization 4, and Visualization 5, and the corresponding velocity fields of GN calculated based on the motion videos are shown in Fig. 4 (b), (d) and (f). These results prove that the massive laser pulling of GN can be achieved in a wide band of visible light, indicating that this phenomenon to a certain degree is universal and robust.

 

Fig. 4. Massive laser pulling of GN with different laser wavelength. (a) Laser pulling images of the GN upon 405 nm laser irradiation. (b) Instantaneous velocity field of the GN in the dotted box area in a. (c) Laser pulling images of the GN upon 532 nm laser irradiation. (d) Instantaneous velocity field of the GN in the dotted box area in c. (e) Laser pulling images of the GN upon 650 nm laser irradiation. (f) Instantaneous velocity field of the GN in the dotted box area in e. The scale bars in a, c and e are 0.4 cm.

Download Full Size | PDF

3.4 Massive laser pulling with different laser power

Then, the effect of incident laser power on the laser pulling of GN in water was investigated by using the 488 nm laser light source, in which the laser power changes from 30 mW to 80 mW. The motion video of GN upon laser irradiation for each power condition is recorded and the corresponding average pulling velocity along the pulling direction is calculated. As shown in Fig. 5, the pulling velocity increases 0.46 mm/s to 0.525 mm/s with an increase in laser power, which should be attributed to a more remarkable asymmetry of the nanosheet temperature distribution for a larger laser power. This result means that the massive pulling behavior is controllable by the input laser energy in potential applications.

 

Fig. 5. The relationship between the average pulling velocity of GN and the incident laser power.

Download Full Size | PDF

4. Discussion

This paper focuses on the experimental results of the laser pulling phenomenon on graphene nanosheets. At present, the mechanism of this phenomenon is still puzzling. Even though, we try to make some discussion about the underlying mechanism herein. The laser light source used in this work are common gauss beams with a plane wavefront, and the typical thickness of GN measured is about several nanometers. In this case, a temperature difference of the GN along the light direction appears, and the end of GN near the light source is hotter generally. This laser pulling phenomenon likely originates from the light-induced thermal effect. In fact, light-induced motion of small object in liquid originated from the asymmetrical temperature distribution has been reported in previous literatures, which is generally called the opto-thermophoresis effect [26]. The velocity of object (u) under a temperature gradient field is given by u =−D_T▽T = −S_T D▽T, where▽T is the temperature gradient, S_T is the Soret coefficient, D_T is the thermophoretic mobility, and D is the Brownian diffusion coefficient (D > 0). Generally, the Soret coefficient S_T = D_T/D is a criterion of thermophoresis [26]. When S_T > 0, the object moves from hot to cold. When S_T < 0, the object moves from cold to hot. In our experiment, if the opto-thermophoresis effect dominates the laser pulling phenomenon, the Soret coefficient S_T should have a negative value. According to the previous researches, the sign and magnitude of the Soret coefficient S_T is sensitive, which can be altered by environmental temperature, surface chemistry of objects, surface charge, etc [26–28]. Sun and Lin et al. detailedly discussed the sign of S_T in their review article [29], in which they also expatiated an interface-force-induced thermophilic behavior (namely S_T = D_T/D < 0) for a charged particle. In terms of this thermophilic behavior, a charged solid−liquid interface with a temperature gradient result in a permittivity (ε) gradient of the water by distorting the structure order of the interface water layer. In this case, according to Anderson’s model, the thermophoretic mobility [29,30]:

5. Conclusions

As a charming light manipulation concept, optical pulling develops rapidly in the past decade. However, optical pulling of graphene in liquid is rarely reported. In this work, massive laser pulling of graphene nanosheets (GN) in pure water is presented. This phenomenon is observed and confirmed by a motion monitoring system composed of a microscope and a CCD device. The velocity field of the GN upon laser irradiation is quantitatively analyzed by using a particle image velocimetry software PIVlab. The massive laser pulling of GN has a certain universality for a wavelength range from 405 nm to 650 nm. The average pulling velocity is about 0.5 mm/s that can be regulated by the incident laser power. The authors hypothesize that this laser pulling phenomenon probably originates from the opto-thermophoresis effect based on the laser-induced asymmetrical temperature distribution of GN. This work supplies a potential massive manipulation technique of small objects in micro-nanofluidic devices and also provides a platform to investigate laser-graphene interactions in a simple liquid phase medium.