Article pubs.acs.org/Langmuir
Cation-Dependent Emergence of Regular Motion of a Float Daisuke Yasui, Hirofumi Yamashita, Daigo Yamamoto, and Akihisa Shioi* Department of Chemical Engineering and Materials Science, Doshisha University, 1-3 Tatara Miyakodani, Kyoto 610-0321, Japan W Web-Enhanced Feature *
ABSTRACT: We report a unique ion-dependent motion of a float at an oil/water interface. The type of motion depended on the cation species that was dissolved in the water. Irregular vibrations occurred when the water contained Ca2+, back-andforth motion occurred when the water contained Fe2+, a type of motion intermediate between these occurred when the water contained Mn2+, and intermittent long-distance travel occurred when the water contained Fe3+. This is one of the simplest systems that can be used to show how macroscopic regular motion emerges depending on specific chemicals, which is one of the central issues in the study of biological and biomimetic motions.
■
INTRODUCTION Moving colloids driven by chemical reactions can move in characteristic ways that resemble the movements of living matter.1−4 Typical examples of this are systems that exhibit chemotactic and stimulus-responsive behaviors. Self-moving colloids containing liquid droplets and catalytic particles have recently been studied in depth.1,3−15 Moreover, autonomously moving vesicles are also examples of self-moving systems. The vesicles can generate rhythmic changes in shape under a chemical potential gradient, and this allows them to actively transport an object.16 Self-moving colloids may be used as carriers and power sources in microfluidic systems17 and could be used in medical applications such as drug delivery. Selfmoving colloids may also be used as model systems to help improve our understanding of the mechanisms involved in a number of types of biological motion.18,19 Biological systems that exhibit self-motion are very complicated. Thus, the simple physics involved in self-motion is likely to be obscured by the complexity of the whole system, even if simple physics governs the motion. Finding a model system that exhibits self-motion with biomimetic characteristics and studying the mechanisms involved in producing the motion may therefore reveal the simple general mechanisms that underlie biological motion. The force that leads to self-motion is generated at the surface of the colloid because a self-moving colloid is essentially an open system in a nonequilibrium state, with chemical reactions or physicochemical events that supply the energy required to produce self-motion at the surface. Therefore, studying the surface dynamics, especially how regular motions emerge, is one of the most promising ways of studying biomimetic selfmotion. The Marangoni effect is often used as a power source in selfmoving colloids.3−11 A well-known example of self-motion is that of camphor on the surface of water, in which various types of motion occur. The motion of camphor has been found to be © 2015 American Chemical Society
controlled by the physical characteristics of the system, such as the geometries of the water surface and the camphor,3,20 and an understanding of the motion exhibited has been gained by taking into consideration how the surface tension gradient is regulated by the moving camphor and the geometry of the experimental apparatus. The control of this motion by these geometrical and physical factors has been studied in detail An oil/water interface containing an oil-soluble anionic surfactant, bis(2-ethylhexyl) phosphate (DEHP), has been found to exhibit self-agitation21 in the presence of Ca2+ and Fe3+. Here we show that the self-agitation of an oil/water interface, caused by Marangoni instability that resulted from chemical reactions, was rectified using different species of cation. A float at the oil/water interface exhibited back-andforth, intermittent, or vibrational motion depending on the cation species present. Most forms of biological motion are sensitive to the presence of specific cations,22 such as Ca2+, Na+, and K+. These cations affect the macroscopic characteristics of the motion. If ionic (chemical) control of the macroscopic motion can be achieved without using biomolecules, then biomimetic motion can be used in numerous technologies using colloidal systems. Moreover, an understanding of how the macroscopic nature of spontaneous motion is produced depending on the cation species provides insight that could aid in the design of colloidal systems in which biomimetic motion occurs. In this study, interplay between microscopic desorption and the macroscopic flow provides a unique spatiotemporal pattern. Received: April 3, 2015 Revised: September 18, 2015 Published: September 22, 2015 11005
DOI: 10.1021/acs.langmuir.5b03049 Langmuir 2015, 31, 11005−11011
Article
Langmuir
■
Once the interface had stabilized, the motion of the powder spread in the interface was recorded. An experiment involving compression of the interface was also performed. For this experiment, a glass funnel was used, as shown in Figure 1e. An oil/water interface formed to prevent the solutions from falling; this was done by closing the lower outlet of the funnel. Charcoal powder was dispersed in the interface as tracer particles. After ∼15 min, part of the lower aqueous phase was dropped, and the position of the interface descended, as shown in Figure 1e. Then, the interfacial area decreases. If molecules adsorbed in the oil/water interface remains, the adsorbed layer in the interface is compressed. Thus, the compression of flat interface is performed. Motions of tracer particles before and after the compression of the interface were recorded. The compression rate of the interface is expressed by (1/ A0)(A0 − A1)/Δt. Here A0 and A1 are the interfacial areas before and after the compression, respectively. Δt is the time required for the compression. When we use the glass funnel shown in Figure 1e, the area of oil/water interface at height z (A(z)) is given by A(z) = π(z tan θ)2. Thus, the compression rate is calculated by (1/A(z0))·(A(z0) − A(z1))/Δt that is equal to (1/z0)(Δz/Δt){2 − (Δz/z0)}. Here z0 and z1 denote the initial and the final z position of oil/water interface, and Δz = z0 − z1. An aluminum sheet was inserted along the end wall of the rectangular cuvette of Figure 1a to allow the effects of the meniscus shape on the float motion to be determined. The aluminum sheet made the wall hydrophobic and changed the meniscus shape. An experiment using the float was performed using this modified apparatus. All float and powder movements were recorded using a microscope (VW-6000/5000, Keyence Corporation). The movie was analyzed using Image-J software to allow spatiotemporal plots to be drawn and Movie Ruler (Photron) software to allow the position of the float to be measured.
EXPERIMENTAL SECTION
Chemicals. Bis(2-ethylhexyl) phosphoric acid (DEHPA, 97%) and manganese(II) chloride tetrahydrate (98%) were obtained from Sigma-Aldrich. n-Heptane (99%), calcium chloride (99%), iron(II) chloride tetrahydrate (99%), iron(III) chloride hexahydrate (99%), and charcoal were purchased from Wako Pure Chemical Industries. Fully deionized water was used. Methods. An aliquot of DEHPA was dissolved in n-heptane to give a concentration of 100 mM. An electrolyte (CaCl2, FeCl2, FeCl3, or MnCl2) was dissolved in water. The electrolyte concentration was usually 10 mM, but concentrations of 100 mM and 1 M were used in some cases. The aqueous phase was poured into a glass cuvette, followed by the oil phase. Once the interface was stabilized, a thin float (usually 6 mm in diameter) made of aluminum sheet was placed on the oil/water interface. Three types of cuvette were used: a 40 mmlong rectangular parallel piped cuvette (Figure 1a,b), a 100 mm-long
■
RESULTS AND DISCUSSION A surface tension gradient causes liquid flow in the surface because the liquid with the higher surface tension pulls the surrounding liquid. This surface flow is sometimes enhanced by hydrodynamic instability. This is called Marangoni instability. At the present oil/water interface, Marangoni instability occurred in numerous places successively, and each occurrence of instability caused radial tangential flow to occur at the oil/ water interface (see Web-Enhanced Object (WEO) Movie 3). The flow pattern was determined by observing the motion of a float placed at the rectangular interface (Figure 1a,b). The float (with a diameter of 6 mm) moved in an approximately 1D manner along the channel when it was placed within the rectangular interface (10 mm wide). The float moved in a random manner when the interface contained Ca2+ ions (see WEO Movie 1). The spatiotemporal pattern of the moving float, which is explained in Figure 1b, is shown in Figure 2a. The float exhibited irregular vibrations, meaning the Marangoni flow driving the motion of the float was generated randomly. The float motions that were found when Fe3+, Mn2+, and Fe2+ were present at a concentration of 10 mM are shown in Figure 2b−d, respectively (see WEO Movie 1). The float repeatedly moved back and forth for at least 4 min when Fe2+ was present, and a similar but less regular back-and-forth motion was found when Mn2+ was present. In contrast, the float moved intermittently, remaining at the end of the channel making small irregular vibrations, when Fe3+ was present. The speeds at which the float moved when Ca2+ and Fe2+ were present are shown in Figure 2e,f, respectively. The speed of the back-and-forth motion was higher than that of the vibrational motion. We have previously found that self-agitation did not
Figure 1. Experimental apparatus. (a) Bird’s eye view of the apparatus with a rectangular channel, (b) top view of the rectangular channel and a spatiotemporal plot of the float motion, (c) bird’s eye view of the circular channel, (d) rectangular channel with a movable wall, and (e) interface compression experiment.
cuvette (Figure 1d), and a concentric cylinder cuvette (Figure 1c). The dimensions of the apparatus are shown in Figure 1a,c,d. A movable wall made of glass was inserted perpendicular to the interface, as shown in Figure 1d, in experiments using the 100 mm long cuvette. The motion of the aluminum sheet was recorded. A rectangular float was also used for the circular channel experiment because it moved more smoothly in the curved channel. The aqueous phase was poured into a glass container with a diameter of 50 mm, followed by the oil phase. Self-agitation of the interface was then observed. Charcoal powder was dispersed in the oil. 11006
DOI: 10.1021/acs.langmuir.5b03049 Langmuir 2015, 31, 11005−11011
Article
Langmuir
Figure 3. Trajectory of the moving float along the circular channel. The scale is shown by the azimuth and by the circumferential length along the inner periphery of Figure 1c.
powder was therefore dispersed at the interface to act as a tracer. A cylindrical container with a diameter of 50 mm was used because the tracer particles tended to aggregate when a narrow rectangular channel was used. Marangoni flow was observed as radially propagated interfacial flow (see WEO Movie 3) appearing at numerous points (sources) in the oil/ water interface. Three consecutive source points are shown in Figure 4a,b. A source point was generated near the boundary Figure 2. Motion of the float along the length of the rectangular channel. Spatiotemporal plots for the (a) Ca2+, (b) Fe3+, (c) Mn2+, and (d) Fe2+ systems. The speed of the float in the (e) Ca2+ and (f) Fe2+ systems. Distributions of the total distances traveled, with the elementary distance indicated. The cation concentration was 10 mM.
occur when other divalent cations, such as Mg2+, Sr2+, Ba2+, Cu2+, or Co2+, were used.21 The results shown in Figure 2a−f were highly reproducible. We performed the experiment five times (4 min per experiment) and obtained a distribution for the distance traveled in one elementary motion, that is, the motion in one direction until the float changed direction, and the distributions found are shown in Figure 2g. The percentage of total distance traveled in the elementary motion is shown. When Fe2+ was present, >83% of the total distance traveled consisted of elementary movements of >20 mm. In contrast, when Ca2+ was present, >90% of the total distance traveled consisted of elementary movements of
