Micro Particles ... - ACS Publications

Subscriber access provided by University of Sunderland

Communication

Controlling the trajectories of nano-/micro particles using light-actuated Marangoni flow Cunjing Lv, Subramanyan Namboodiri Varanakkottu, Tobias Baier, and Steffen Hardt Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b02814 • Publication Date (Web): 04 Oct 2018 Downloaded from http://pubs.acs.org on October 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Controlling the trajectories of nano-/micro particles using lightactuated Marangoni flow

Cunjing Lva,b,1, Subramanyan Namboodiri Varanakkottuc,1,2, Tobias Baiera, and

Steffen

Hardta,2 aInstitute

for Nano- and Microfluidics, Technische Universität Darmstadt, Alarich-Weiss-Straße 10, 64287 Darmstadt, Germany bDepartment

cDepartment

of Engineering Mechanics, Tsinghua University, 100084 Beijing, China

of Physics, National Institute of Technology Calicut, Kozhikode 673601, Kerala, India 1These 2Correspondence

authors equally contributed to the work.

to: [email protected], [email protected]

Abstract The ability to manipulate small objects and to produce patterns on the nano- and microscale is of great importance, both with respect to fundamentals and technological applications. The manipulation of particles with diameters of the order of 100 nm or below is a challenge because of their Brownian motion, but also because of the scaling behavior of methods such as optical trapping. The unification of optical and hydrodynamic forces is a potential route towards the manipulation of tiny objects. Herein we demonstrate the trapping and manipulation of nano- and microparticles based on interfacial flows controlled by visible light, a method we denote “LightActuated Marangoni Tweezer (LAMT)”. We experimentally study the manipulation of particles having diameters ranging from 20 nm to 10 μm, including quantum dots and polystyrene nano/microparticles. The particles can be manipulated by scanning a light beam along a liquid surface. In this way, we are able to define almost arbitrary particle trajectories, for example in the form of letters. In addition to that, we are able to handle a number of particles in parallel by creating an optical “landscape” consisting of a multitude of laser spots. The inherent advantages of LAMTs are the linear scaling of the trapping force with the particle diameter and the fact that the force is less dependent on particle properties than in the case of conventional methods.

1

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

KEYWORDS: nano-/micro particles, Marangoni stress, optofluidics, surfactant, interfaces, particle trapping Precise control over the positions and trajectories of nano-/microscale objects is crucial for many emerging areas of nanoscience and technology.1–4 Examples where controlling the position or arrangement of small-scale objects is required are force measurements with femtonewton sensitivity,5 spectroscopy of single molecules,6 or tunable optical systems.7,8 Manipulation strategies based on optical forces are most preferred because of the inherent advantages of this technique such as non-contact manipulation, tunability (in terms of wavelength, intensity, polarization and beam profile) and programmability.9 These so-called optical tweezers revolutionized the area of particle manipulation.9,10 Nevertheless, optical tweezers suffer from serious limitations. Specifically, the manipulation of tiny objects (of a size of 100 nm or below) remains a challenge because of the inherent gradient nature of the trapping force which scales with the particle volume and diminishes rapidly when the particle size approaches the submicron scale.1,10 Compensating the unfavorable force scaling by increasing the light intensity is often not a solution, since the danger of damaging the specimen increases, and light absorption causes undesired heating. Furthermore, while optical gradient forces are of short-range nature, it may be desirable to collect or trap objects from a comparatively large volume. Manipulation techniques which utilize light in combination with nanostructured surfaces,11–14 photosensitive solids15 and/or hydrodynamic flows14,16–26 offer a potential way to overcome some of these fundamental challenges. For example, nano-aperture based plasmonic optical tweezers have been employed for the precise trapping of nanoscale objects.11–13 Further, an optically-based method has been demonstrated where an electrothermal flow is induced by local heating of a nanostructured surface due to surface plasmon excitation.14 Although these methods enable nanoparticle trapping and manipulation, they require nanostructured substrates, which means that the particle positions are predefined according to given geometrical features. They do not allow a completely free choice of particle trajectories, as it is in principle possible with classic optical tweezers.

2


Page 2 of 21

Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Methods that utilize hydrodynamic forces to manipulate particles offer specific advantages such as a favorable force scaling (scaling with the particle diameter), independence of optical particle properties, a low light intensity, and a large area of influence.14,16–26 Herein we present a technique for the manipulation of micro- and nanoparticles based on light-induced Marangoni flow termed “Light-Actuated Marangoni Tweezer (LAMT)”. The mechanism relies on light-induced surfacetension driven flow due to the physicochemical changes of a photosensitive liquid surface. This work is based on a previous study22 in which we demonstrated the manipulation of single 15 µm polystyrene spheres at a photosensitive liquid surface using UV light. Employing UV light for particle manipulation may limit the applicability of this method because of the potential damage to the specimen caused by UV light. In this work we overcome this limitation by demonstrating LAMT-based particle manipulation based on visible light (442 nm). Moreover, we show the applicability of this method in the nanoworld, and that it holds the potential for highly parallel manipulation of small-scale objects. We experimentally demonstrate the manipulation of particles having sizes ranging from 20 nm to 10 μm, including quantum dots and polystyrene nano/microspheres. Scanning a laser spot along the liquid surface allows guiding these objects along complex trajectories. In addition to single particle manipulation, we have established the parallel manipulation of particles by creating an optical landscape consisting of multiple laser spots. Working principle and experimental setup The LAMT principle is shown in Figure 1A and is based on the photoswitching of surfactants between two isomeric states, corresponding to different surface-tension values.

3


Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. Working principle and experimental setup. (A) Schematic of the LAMT working principle. (B) Schematic of parallel manipulation of two particles with two laser spots. (C) Experimental setup including a diffractive optical element (DOE) and a tiltable, piezo-controlled mirror (M2). (D) Microscopy images of light patterns with two and four spots created at the liquid surface.

The photosurfactant diethylene glycol mono(4’,4-butyloxy,butyl-azobenzene) (C4AzoOC4E2) used in the experiments reported here exists in two isomeric states: a trans and a cis state.27 When the surfactant is dissolved in water, at a concentration above the critical micelle concentration (CMC), the water-air interface is mostly covered with molecules in their trans state. The CMC of the trans state is 1.6 µM, while that of the cis state is 23.8 µM.27 The trans-cis switching of the photosurfactant solution can be triggered either using UV or blue light illumination, depending on the photostationary state of the solution22,24,28 (Figures S1–S3, Supporting Information (SI)). We worked with the blue-adapted photostationary state of the surfactant solution (100 µM and 250 µM), which contains both the trans and the cis isomers in the bulk (about 86 % trans and 14 % cis28,29). The photostationary state of the surfactant solution is obtained by a 30 min exposure of the dark-adapted solution under blue light and a subsequent 10 hour period in the dark. After that, the surface is covered almost exclusively by trans isomers. In the experiments reported here, we only used 442 nm irradiation, mainly for the reason that light of 325 nm wavelength could be harmful to specimens sensitive to UV light. The trans-cis isomerization of surfactant molecules adsorbed at the surface results in an increase in surface tension of approximately 10 mN·m–1.28 The 4


Page 4 of 21

Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

difference in surface tension between the irradiated and non-irradiated area results in a surfacetension driven flow,22 known as Marangoni flow30 directed towards the region with higher surface tension (i.e. towards the laser spot). This flow exerts a hydrodynamic force onto a particle attached to the liquid surface, allowing to trap the particle or drag it along the surface when the laser spot is moving. In order to promote the attachment of the particles to the liquid-air interface, we add a small amount of polyethylene glycol (PEG) to the photosurfactant solution (more details are given in Table S1, SI). Figure 1B shows a schematic of two laser spots and two particles attracted to the spots. Figure 1C is a sketch of the experimental setup. The 442 nm emission of the He-Cd laser (Kimmon Koha, Japan) passes through a diffractive optical element (DOE) to create a light pattern with multiple laser spots. The light pattern is focused on the photosensitive liquid surface. Alternatively, without DOE only a single laser spot is created. In order to change the position of the light pattern on the surface, the mirror M2 is attached to a piezo tilting system (PI, model E-500.00X, Germany). Hence M2 can be tilted using preset parameters in the control software or by a joystick. Figure 1D shows microscopy images of light patterns with two and four spots produced by the DOE. More details on the experimental setup are given in the SI, section 1.1. In the experiments we used a 1 mm thick layer of photosurfactant solution in a Petri dish of 6 cm diameter. Light absorption by the 1 mm thick liquid layer could result in two effects: solutal Marangoni flow as discussed, and thermal Marangoni flow due to the conversion of electromagnetic energy into heat. To estimate an upper bound for the surface tension variation due to temperature increase, we assume that 100 % of the input power is converted into heat which is transported to the perimeter of the Petri dish by conduction. This very conservative estimate yields a temperature rise of ~ 2.5 K at the laser spot,31 corresponding to a surface tension decrease of ~ 0.4 mN·m‒1,32,33 while photoswitching leads to an increase in the surface tension of ~ 10 mN·m‒1.27 This suggests that the heating effect resulting from the laser source has a negligible influence on the Marangoni stresses created at the liquid surface (details of this analysis are given in the SI, section 1.3). Particle manipulation with multiple laser spots 5


Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

First, we demonstrate the effects due to multiple laser spots and the potential of parallel particle manipulation. Owing to the comparatively low power required to trap or drag a particle, our method is well suited for such operations. In each of the experiments, a total power of 2 mW was equally distributed over multiple laser spots. In the first experiment, two spots with an approximate diameter of 15 µm were created, separated by a distance of approx. 170 µm. Single particles (polystyrene spheres, 10 µm in diameter) were carefully deposited at the liquid surface close to the irradiated area.

Figure 2. Particles affected by multiple laser spots. (A, B) Distance travelled as a function of exposure time in an arrangement of two laser spots for particles initially at positions A, B, C and D. The illumination starts at t = 0. The insets show the corresponding particle streak velocimetry images. (C) Time lapse showing the motion of four particles induced by four laser spots. The final positions are reached at t = 20 s.

To study the effect of the light pattern on a particle, we evaluated the trajectory of particles initially at positions A, B, C, and D, as indicated in Figure 2A,B. As shown in Figure 2A, after the laser has been switched on, the particles initially at A, B and C go to the nearest laser spot (Movie S1) 6


Page 6 of 21

Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

where they get trapped. In order to quantitatively characterize their dynamic behavior, particle streak velocimetry measurements were carried out (using the NIS-Elements Software), by which the relationship between the particle position d and time t can be obtained. Velocities of the order of 20 µm·s‒1 are reached, and a decrease in velocity is observed when a particle comes close to a laser spot. This is presumably due to the crowding of surfactant molecules induced by the inward flow along the surface.22 Three characteristic particle trajectories overlaid in one image are shown in the inset of Figure 2A. For a particle being initially at an equal distance to the two laser spots, the two force fields compensate each other as soon as the particle has reached the straight line connecting the two spots. This is demonstrated in Figure 2B. In that case, the distance shown at the y-axis is defined as d(t) = [d1(t) + d2(t)]/2, where d1(t) and d2(t) are the distances between the particle center and the centers of the laser spots 1 and 2, respectively. The particle stops as soon as it has reached the straight line connecting the two laser spots (see also Movie S1 and Figure S4). In Figure 2C, we show the results of parallel manipulation of four particles with four laser spots. In that case, the ensemble consists of three individual particles and an agglomerate of two particles (the one initially closest to spot #1). We used a DOE to create four identical laser spots, each having a diameter of approx. 18 µm, with a diagonal separation of approx. 180 µm. Figure 2C displays the time lapse images of the particles being attracted by the spots. It is observed that the final positions of the particles are situated slightly inside the region enclosed by the four spots. This behavior can be attributed to the long-range nature of the flow fields. To qualitatively explain how the flow field due to multiple lasers spots is related to the flow field of single spot, it is first useful to note that the Reynolds number in our experiments is significantly smaller than one. This means that the Navier-Stokes equation reduces to the Stokes equation, meaning that the linear superposition of individual flow fields represents a solution of the momentum equation. However, when trying to apply this principle to our experimental situation, the intricate coupling between the flow field and the surfactant distribution needs to be considered. Strictly, the fact that the surfactant distribution around a laser spot also depends on the flow field due to a neighboring spot invalidates the superposition principle. Since assessing the importance of such complex interaction effects between neighboring spots is very difficult, in the following 7


Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

we make the simplifying assumption that the superposition principle holds. Correspondingly, at a specific position the net flow velocity is given by the superposition of the four individual flow fields. As a consequence, the positions where the flow velocity vanishes are not identical to the positions of the laser spots. This can be shown by referring to the experimentally determined velocity field around a single laser spot (see Figure S5, SI). The examples of particles exposed to two and four laser spots demonstrate that our method holds the potential to manipulate many particles in parallel, but that the resulting flow field is complex due to the superposition of different long-range contributions of the individual spots. In comparison to holographic trapping,34 LAMTs have the advantage that due to the low power requirements it becomes much easier to practically implement optical landscapes with a large number of individual traps. Trapping and dragging of nanoparticles One of the major aims of this work is to realize light-controlled manipulation of nanoscale objects. We expect that the particles at the interface experience the same flow field, irrespective of their size. To study the influence of the particle size, we performed experiments with polystyrene spheres having diameters of 10 μm, 1 μm and 100 nm. The bigger particles (i.e. 10 μm and 1 μm) were deposited directly at the liquid surface from dry powers using a small scoop. Since commercial nanoparticles are already in suspension and cannot be deposited in the same way, we carefully placed a tiny drop (e.g. ~ 1 μL) of nanoparticle suspension with a very low particle concentration at the liquid surface. That way we were able to detect individual nanoparticles at or near the liquid surface. We considered particles initially at a distance of approx. 150 μm from the laser spot, and evaluated the resulting particle trajectories. To achieve this, a 442 nm beam with a power of 1.1 mW was focused at the interface to a single spot of 11 µm diameter.

Figure 3 shows the relationship between the distance travelled and time, for all the particles considered here. A streakline of a 100 nm particle is shown in the inset of Figure 3. Notably, the average velocity with which the particles are attracted to the trap decreases with particle diameter. 8


Page 8 of 21

Page 9 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Particles with a diameter of 10 μm exhibit an average velocity of about 7 µm·s‒1, which decreases to about 1 µm·s‒1 for 100 nm particles. To uncover the reasons behind this behavior, we studied the influence of Brownian motion (see SI, section 1.5). From evaluating the Brownian motion of 100 nm particles at the water-air interface we obtain a diffusion coefficient of D = 1.6 μm2·s–1. It takes about 150 s for the particles to travel the distance from their original position to the laser spot. Simply by diffusion the particles would have traveled a distance of about 30 µm during the same time span. Based on the fact that the advective and the diffusive displacement of 100 nm particles are of the same order of magnitude, we conclude that Brownian motion is a potential reason for the different trajectories visible in Figure 3. Due to the scaling of the diffusion coefficient with the inverse particle diameter, Brownian motion influences the larger particles to a lesser extent. Moreover, since the commercial nanoparticle suspension contains some surfactant as a stabilizing agent, it cannot be ruled out that traces of this surfactant are responsible for the lower velocity of the smaller particles compared with the bigger ones. A second surfactant displacing the photosurfactant from the liquid surface could reduce the driving force for the flow.

Figure 3. Micro- and nanoparticle trapping. Distance traveled as a function of time of 10 µm, 1 µm and 100 nm polystyrene particles initially 150 µm away from a laser spot. The inset shows a streakline of a 100 nm particle.

The LAMT allows controlling the trajectories of nanoparticles over large distances. To demonstrate that, we scanned a 1.1 mW laser beam along the liquid surface over a distance of 9


Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

about 100 μm with a velocity of a few μm·s–1. In Figure 4A, we show the corresponding manipulation of a 100 nm polystyrene particle. After switching on the laser beam, the particle moved towards the laser spot and stopped very close to it. From that moment on, the mirror M2 was tilted using a joystick to move the laser spot along a straight line. In Figure 4A it can be seen that the nanoparticle closely follows the laser spot when it is translated with a velocity of approx. 1 μm·s–1. The rightmost frame in that figure is a long-time exposure showing the particle trajectory as an almost straight line. The blurred band superposed to that line indicates the surface regions having been exposed to laser light. Figure 4B demonstrates the relationship between the distance d traveled by the particle and time t. The maximum velocity with which we can drag a 100 nm polystyrene particle using the 1.1 mW beam is ~ 7 μm·s–1, above which the particle escapes from the trap.

Figure 4. Manipulation of nanoparticles. (A) Time lapse of a single 100 nm fluorescent polystyrene particle dragged along a straight line. The distance travelled as a function of time is shown in B. (C, D) show analogous results obtained for a 20 nm quantum dot.

A similar operation was performed with a 20 nm quantum dot. Corresponding results are shown in Figure 4C,D. Compared to the 100 nm particles, Brownian motion is much more pronounced, which results in fluctuations around the straight trajectory defined by the laser (Figure 4D). The blurred band superposed to the particle streak indicates the surface regions having been exposed to laser light. A maximum transport velocity of 4 μm·s–1 is achieved here, which is lower than that

10


Page 10 of 21

Page 11 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

of the 100 nm particles under similar irradiation conditions. This reduction in velocity is attributed to the increased Brownian motion of the quantum dots at the interface, as evident from Figure 4D The LAMT method also allows dragging particles along complex trajectories extending over comparatively large areas of the liquid surface. Figure 5A shows trajectories of 100 nm particles in the form of the letters “C”, “S”, and “I”. Owing to Brownian motion, manipulating 20 nm quantum dots is significantly more difficult. The result of an attempt to “write” the letter “O” is shown in Figure 5B. Along some sections of the trajectory the quantum dot is not visible, presumably because it has diffused away from the liquid surface. To obtain the trajectories shown in Figure 5, the laser spot was moved along the surface with velocities ranging between 2 and 7 µm·s‒1.

Figure 5. Dragging nanoparticles along complex trajectories. (A) Trajectories of 100 nm fluorescent polystyrene particles forming the letters “C”, “S” and “I”, standing for “Center of Smart Interfaces”, a research center at the TU Darmstadt. (B) Trajectory of a 20 nm quantum dot, forming the letter “O”.

The LAMT principle offers the advantage of particle manipulation at comparatively low light intensities, as discussed in the following. A simple estimate of the Stokes drag force reveals that a force of a few fN is applied on the nanoparticles in the velocity range of 1 – 7 μm·s–1. To exert optical gradient forces of a few fN, in the case of Rayleigh scattering (r

Micro Particles ... - ACS Publications

Recommend Documents