Subscriber access provided by UNIVERSITY OF CONNECTICUT
Article
Active Janus Particles at Interfaces of Liquid Crystals Rahul Mangal, Karthik Nayani, Young-Ki Kim, Emre Bukusoglu, Ubaldo M. Córdova-Figueroa, and Nicholas L. Abbott Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b02246 • Publication Date (Web): 29 Aug 2017 Downloaded from http://pubs.acs.org on September 2, 2017
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 free 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 accessible to all readers and 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.
Langmuir 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 12
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
Langmuir
Active Janus Particles at Interfaces of Liquid Crystals Rahul Mangal,a Karthik Nayani,a Young‐Ki Kim,a Emre Bukusoglu,b Ubaldo M. Córdova‐ Figueroa,c and Nicholas. L. Abbott a* a
Department of Chemical and Biological Engineering, University of Wisconsin‐Madison, 1415 Engineering Drive, Madison, Wisconsin 53706, United States Chemical Engineering Department, Middle East Technical University, Ankara, 06800 Turkey
b c
Department of Chemical Engineering, University of Puerto Rico – Mayagüez, Mayagüez, PR 00681, USA
ABSTRACT: We report an investigation of the active motion of silica‐palladium Janus particles (JPs) adsorbed at interfaces formed between nematic liquid crystals (LCs) and aqueous phases containing hydrogen peroxide (H2O2). In comparison to isotropic oil‐aqueous interfaces, we observe the elasticity and anisotropic viscosity of the nematic phase to change qualita‐ tively the active motion of the JPs at the LC interfaces. Although contact line pinning on the surface of the JPs is observed to restrict out‐of‐plane rotational diffusion of the JPs at LC interfaces, orientational anchoring of nematic LCs on the silica (planar) and palladium (homeotropic) hemispheres bias JP in‐plane orientations to generate active motion almost exclu‐ sively along the director of the LC at low concentrations of H2O2 (0.5wt%). In contrast, displacements perpendicular to the director exhibit the characteristics of Brownian diffusion. At higher concentrations of H2O2 (1wt % ‐ 3wt%), we observe an increasing population of JPs propelled parallel and perpendicular to the LC director in a manner consistent with active motion. In addition, under these conditions, we also observe a subpopulation of JPs (approximately 10%) that exhibit active motion exclusively perpendicular to the LC director. These results are discussed in light of independent measurements of the distribution of azimuthal orientations of the JPs at the LC interfaces and calculations of the elastic energies that bias JP orientations. We also contrast our observations at LC interfaces to past studies of self‐propulsion of particles within and at the interfaces of isotropic liquids.
INTRODUCTION Active colloidal particles (living and synthetic) exhibit dy‐ namic behaviors that are distinct from thermal fluctuation‐ induced Brownian motion as they are self‐propelled by net forces generated at the particle level.1‐8 For example, in bi‐ ological systems, bacteria and unicellular protozoa use their flagella or cilia for active swimming.1,2 For synthetic colloids, self‐propulsion can be achieved by converting ex‐ ternal energy (chemical or field), via the creation of local gradients (chemical or physical), into net forces.3‐13 The origin of these gradients is often related to asymmetry in structure at the single particle level. For instance, spherical Janus (two‐faced) particles (JPs) with a single Pt/Pd‐coated face (JPs) and bimetallic rods have been shown to display active motion in aqueous H2O2 solutions.1–6 The Pt/Pd sites act as catalysts for decomposition of H2O2 and the associ‐ ated concentration gradients lead to self‐diffusiophoresis. Alternatively, when JPs are illuminated with a laser beam, selective heating of a metal patch can also drive the motion of the particles via self‐thermophoresis.7–9 In addition to exhibiting rich physics, active colloidal systems have po‐ tential utility in contexts such as drug delivery, macromol‐ ecule separations, and colloidal self‐assembly.10–13 At fluid interfaces, active motion of JPs is strongly influ‐ enced by wetting of the particle surfaces by the fluids, and the coupling of this wetting to the distribution of orienta‐ tion of the JPs.14–18 Relevant phenomena include pinning of contact lines14 and associated capillary interactions.15,19 Re‐ cently, Wang et al.16 and Malgaretti et al.17 investigated via experiment and theory, respectively, the motion of cata‐ lytic JPs at the air–water interface and found that the re‐ stricted rotation of the JPs (due to pinning of contact lines)
can lead to enhanced activation at the interface as com‐ pared to bulk. Dominguez et al. have also discussed Ma‐ rangoni stresses that can be generated by catalytic particles at an interface, leading to active motions.18 The majority of past studies of active particles have focused on Newtonian fluids within which the hydrodynamics gov‐ erning colloidal dynamics are well understood.1–6,16–18,20 An overarching concept emerging from these studies is that active particles in Newtonian fluids show sustained motion in a given direction on time‐scales that are short compared with the characteristic rotational time‐scale. In contrast, on time scales that are long compared with the rotational time‐scale, the active particles undergo random walks with an enhanced effective diffusion coefficient (as compared to passive diffusion). Additionally, in Newtonian fluids, the rotational motion of particles is decoupled from processes that give rise to propulsive forces.4 For non‐Newtonian flu‐ ids, however, reciprocal coupling of colloidal dynamics with the rheological and structural properties of the fluid results in rich and distinct phenomenology. For example, recent theoretical predictions suggest that fluid elasticity can either enhance or retard colloidal propulsion.21–26 Mo‐ tile microorganisms such as E. coli have been observed to swim in highly viscous gel‐forming fluids27 or elastic me‐ dia28 at velocities that exceed that observed in water. Re‐ cently, Paulo and co‐workers29,30 propelled magnetic parti‐ cles via reciprocal stroke swimming in polymer solutions, a mechanism of swimming that is prevented in Newtonian fluids due to flow reversibility at low Reynolds number. In this paper, we consider active particles in contact with liquid crystals (LCs), as an example of an anisotropic, non‐ Newtonian fluid. LCs are complex fluids that exhibit long‐
ACS Paragon Plus Environment
Langmuir
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
range orientational order, resulting in anisotropic optical and mechanical properties.31 The elasticity of these fluids results in a resistance to deformation from a preferred ori‐ entation, the so‐called director no.31,32 The elasticity and anisotropic viscosities of LCs have been documented to in‐ fluence passive diffusion of colloids,33,34 leading to aniso‐ tropic trajectories and preferred orientations (for non‐ spherical particles).35 Of particular relevance to the investi‐ gation reported in this paper, recent studies by us and oth‐ ers of bacterial motion in bulk nematic lyotropic LCs reveal a rich range of active behaviors, including director‐guided motion36,37 and dynamic self‐assembly.38 These studies also revealed that, at isotropic‐nematic LC interfaces, the ne‐ matic elasticity and presence of topological defects can be used to capture and release motile bacteria.39 Other studies have reported collective bacterial behaviors that are medi‐ ated by the anisotropic elasticity of LCs.40 Overall, these past observations highlight the potential utility of LCs as structured fluids to direct the motion of synthetic active colloids to a specific destination.41–45 Herein we report on the dynamics of active JPs in contact with LC interfaces. Specifically, we leverage past studies of self‐diffusiophoresis of metal‐coated JPs in Newtonian flu‐ ids3,4,6 to elucidate the physics of active colloidal behaviors in the anisotropic environment of LC‐water interfaces. We address several key questions: (i) How does the interplay of LC‐mediated elastic interactions, interfacial energies and contact line pinning influence the orientations of JPs at the interface? (ii) How does the dynamic distribution of orientations of JPs influence the active motion of JPs at the LC aqueous interface? (iii) How do the transport proper‐ ties of active particles at LC interfaces differ from those ob‐ served at isotropic fluid interfaces?
EXPERIMENTAL We used 4‐pentyl‐4’‐cyanobiphenyl (5CB), a thermotropic LC that exhibits a nematic phase between 23.7C and 35.3C.31 JPs were synthesized following the procedure de‐ scribed by Love et al.46 Briefly, a monolayer of silica micro‐ particles (diameter (d) of 3.03±0.17 µm; microParticles GmbH) was deposited on a Piranha‐cleaned glass substrate by drop casting. Subsequently, using a thermal evaporator, a 3nm‐thick layer of Cr and a 10 nm‐thick layer of Pd were sequentially deposited on the monolayer of particles. The JPs were then dispersed into de‐ionized (DI) water by son‐ ication for approximately 5 minutes. Dilute suspensions of the JPs were cleaned by repeated centrifugation. The LC‐water interfaces used in our study were prepared by dispensing a small amount of 5CB (nematic phase) into the pores of a gold‐coated transmission electron micros‐ copy (TEM) grid (20 µm thickness). The grid was sup‐ ported on a polyimide‐2555 (PI)‐coated indium‐tin‐oxide (ITO) glass substrate, where the PI was used to enforce tan‐ gential or planar anchoring of nematic 5CB. The lateral di‐ mensions of each individual pore of the TEM grid was ap‐ proximately 284 µm X 284 µm, much larger than the parti‐
Page 2 of 12
cle diameter (3.03±0.17 µm). Before use, the PI‐coated sub‐ strate was rubbed to achieve unidirectional alignment of contacting nematic phases. Subsequently, the glass sub‐ strate supporting the grid filled with LC was carefully im‐ mersed into a dish containing DI water to form the LC‐ water interface. A small droplet of an aqueous dispersion of JPs was introduced into the aqueous phase overlying the LC‐filled TEM grid. An aliquot of a 100 mM solution of NaCl was then added to the aqueous phase to screen the electrical double layer repulsion between the LC interface and the sedimenting JPs (the final salt concentration in the aqueous phase was 10 mM). After 10 minutes, the salt was diluted to trace concentrations (see Supporting Infor‐ mation (SI) for additional details). Finally, a calculated vol‐ ume of a 30 wt% H2O2 aqueous solution was added to the dish to bring the final concentration of H2O2 in the aque‐ ous phase contacting the LC to the desired value (0–3 wt%). The dish was then sealed with a coverslip. We also performed experiments at H2O2 concentrations higher than 3 wt% but observed the formation and coalescence of bubbles to generate bulk convection that impacted the mo‐ tion of JPs. Additionally, the depth of the water sub‐phase was limited to 800 µm to minimize convection. We meas‐ ured the nematic‐to‐isotropic transition temperature of 5CB to be unchanged by incubation with 3 wt% H2O2 for 4hrs. For control measurements, an isotropic oil (silicone oil S159500, oil 4.8 cP, oil 0.97 g/ml) was used instead of 5CB. Although the oil is slightly lighter than water, ca‐ pillary forces trap the oil within the TEM grid under water. We imaged JPs at the LC interface in bright‐field mode us‐ ing an Olympus BX60 microscope equipped with cross po‐ larizers. Displacements of isolated JPs at the 5CB–aqueous interface were recorded over a period of 4‐5 minutes, at 4 frames per second. A monochrome camera (Cohu4915‐ 2010/0000) with a resolution of 640 X 480 pixels was used. Particle tracking was performed with Image‐J using an im‐ age correlation‐based approach to obtain particle trajecto‐ ries [mean square displacements (MSD (X(µm), Y(µm)) vs time interval t (s)]. Still micrographs of JPs at the interface were also captured. To characterize the orientations of JPs in bulk nematic 5CB and also the anchoring of LCs on the surfaces of the JPs, JPs with diameters of 5 m (synthesized using the same proce‐ dure as described above) were dispersed in nematic 5CB. A small volume of 5CB containing JPs was injected between two glass substrates (surface separated by 20 m) with surfaces coated with rubbed PI (anti‐parallel orientation). Micrographs, both in parallel and cross‐polarized configu‐ rations, were captured.
RESULTS
ACS Paragon Plus Environment
Page 3 of 12
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
Langmuir tion of the elastic energy of the LC) and aligns the LC uni‐ formly at the LC‐water interface,47 which would otherwise exhibit degenerate planar anchoring. Figure 1(e) shows a bright field optical micrograph of a sin‐ gle LC‐filled square of a TEM grid onto which JPs were sed‐ imented. The highlighted (dotted circles) dark spots are the JPs adsorbed at the LC‐water interface. Past studies have established that adsorption energies of micrometer‐ sized particles at aqueous‐5CB interfaces are at least 107 kBT.48 Accordingly, microparticles adsorbed at aqueous‐LC interfaces do not escape the interface to either the bulk wa‐ ter or LC phase by thermal fluctuations (kBT) and, in‐ stead, move in the 2D interfacial plane (JPs trapped at the LC‐water interface remained within the focal plane). Ad‐ dition of salt, however, was found to be necessary to screen the electrical double layer repulsion between the sedi‐ menting JPs and the 5CB‐water interface in order to achieve the adsorbed state. In the absence of the salt, we observed the JPs to levitate above the LC interface and not penetrate into it. For the reason stated above, however, once adsorbed, the JPs do not desorb upon lowering of the ionic strength of the aqueous phase.
Figure 1: (a) Bright‐field micrographs of Pd‐SiO2 JPs (b‐c) Schematic of experimental setup. R indicates the rubbing direction. (d) Cross polarized micrographs of TEM grid filled with 5CB submerged under water. P and A indicate the polarization of polarizer and analyzer. (e) Bright‐field micrograph of JPs adsorbed at LC‐water interface.
Figure 1(a) shows four bright‐field micrographs of JPs sup‐ ported on glass substrates, displaying distinct silica (bright) and Pd (dark) regions. In Figure 1(a)(iii), the JP appears uniformly dark, which corresponds to a JP oriented such that the Pd patch only is visible from above (viewing direction). In contrast, JPs oriented with the SiO2 surface upward had a bright optical appearance. Figure 1(b‐c) shows the experimental set‐up used to prepare LC inter‐ faces decorated with JPs. Polarized optical micrographs of TEM grids filled with LCs and immersed under water confirmed the uniform orien‐ tation of the LCs at the aqueous interfaces. Specifically, in‐ spection of Figure 1(d) reveals that the director (no) is aligned in an azimuthal direction that is parallel to the di‐ rection (R) in which the PI film supporting the LC was rubbed. The unidirectional alignment of the LC on the rubbed PI propagates through the LC film (via minimiza‐
To validate our experimental procedures, we first per‐ formed measurements of the mean‐squared displacements (MSD) of JPs adsorbed at the interface of isotropic oils and water, in the absence of H2O2. For this control experiment, MSD(X,Y) was confirmed to be a linear function of time, as shown by the representative MSD vs t plot (Figure 2(a)) for a particular JP, consistent with Brownian motion. In ad‐ dition, the absence of anisotropy in the values of the MSDs measured in orthogonal directions on the interface con‐ firmed the absence of bulk convection. Similar measure‐ ments for approximately 20 other JPs were performed re‐ sulting in DX=0.011±0.005 µm2/s and DY=0.011±0.004 µm2/s. Here we note that the values of the diffusion coefficients are intermediate to Doil=0.003 m2/s and Dwater = 0.16m2/s, where Doil and Dwater are diffusion coefficients calculated using the Stokes‐Einstein equation (oil=36.5 cP and water =0.9 cP) for JPs in bulk oil or water, respectively. We also measured MSD(X,Y) for JPs trapped at the nematic LC‐aqueous interface in the absence of H2O2. As shown in Figure 2(b), the MSDs were linear with time, consistent with simple diffusive behavior. However, MSD(Y) exhib‐ ited a higher slope as compared to MSD(X), yielding to DY= 0.018±0.001 µm2/s and DX= 0.010±0.001 µm2/s. The anisot‐ ropy in diffusion coefficients (DY/DX > 1) results from the orientationally ordered nematic LC sub‐phase which al‐ lows preferential displacement of JPs along the far‐field di‐ rector (no Y).33,34,48 Similar measurements for approxi‐ mately 60 other JPs were performed, resulting in DY =0.024±0.017 µm2/s, DX=0.012±0.009 µm2/s and thus DY/DX=2.0±0.7. For comparison, in bulk 5CB, we calculate D5CB (0.0038 m2/s)
