Template-Assisted GLAD: Approach to Single and Multipatch Patchy

Dec 6, 2013 - Template-assisted glancing angle deposition (GLAD) is explored for the fabrication of single and multipatch patchy particles with one or...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/Langmuir

Template-Assisted GLAD: Approach to Single and Multipatch Patchy Particles with Controlled Patch Shape Zhenping He† and Ilona Kretzschmar*,†,‡ †

Chemistry Department, The Graduate Center, The City University of New York, 365 Fifth Avenue, New York, New York 10016, United States ‡ Department of Chemical Engineering, The City College of New York, 140th Street & Convent Avenue, New York, New York 10031, United States S Supporting Information *

ABSTRACT: Template-assisted glancing angle deposition (GLAD) is explored for the fabrication of single and multipatch patchy particles with one or more patches of controlled but asymmetric shape. The template is used to ensure the formation of uniform patchy particles, whereas rotation of the template gives access to a large number of asymmetric patch shapes caused by the shadowing effect of the templating groove and/or the neighboring particle. Careful analysis with a straightforward geometric model reveals the effect of the angle of incidence, rotational angle, groove size, and particle size on the patch shape. Initial magnetic field assembly results are presented to illustrate the removal of patchy particles from their template and accessibility to a large number of patchy particles. Two-patch particles with overlapping patches are also accessible by means of secondary metal vapor deposition. The connectivity of these two patches on each particle and the predictable size of the overlapping section provide access to functional patchy particles. The combination of the template-assisted GLAD method with rotation of the template and secondary evaporation is demonstrated to be a good method for fabricating patchy particles with a variety of asymmetric patch shapes, sizes, and multipatches where every particle of a batch carries exactly the same patch pattern and thereby provides valuable input on experimentally accessible patch shapes for future experimental and computational studies of patchy particles.

■

quality of the mask packing pattern, which needs fine control to be obtained. In light of bulk fabrication, the throughput of this method is rather limited. The glancing angle deposition method (GLAD) is successful in achieving various patch shapes by changing the angle of incidence and monolayer orientation.19,20 However, the diversity of domain orientations in monolayers causes a mixture of various patch shapes in each run, resulting in a lack of patch uniformity necessary to achieve specific assembly structures. It also causes difficulty in increasing the productivity of the method and in separating particles with differing patches. Inspired by the method of fabricating metal shells using GLAD, nanometer- and micrometer-sized metal structures of specific shapes have been made21 that can assemble into specific shaperelated patterns and have shape-dependent properties with potential applications in optics and electronics.22 Most recently, Shah et al.23 reported the preparation of spheroidal Janus and kayak-shaped patchy particles via elongational stretching under heating before or after metal deposition, respectively, and investigated their bulk assembly behavior at various salt concentrations. However, how the patch symmetry affects the assembly structures has not been investigated experimentally, in

INTRODUCTION The preparation and assembly of patchy colloidal particles has received major attention in recent years owing to the tunable interaction potentials and unique properties of patchy colloids.1−9 Complex structures such as multipods assembled from Janus colloidal matchsticks,4 a colloidal Kagome lattice formed from two-pole patchy particles,10 polymer chains resulting from bifunctional patchy particles,11 doublets obtained from patchy colloidal particles,12,13 and even colloidal analogues of molecules assembled from patchy particles14 have been reported. However, all of these assemblies are based on patchy colloidal particles with symmetric patches (in most cases circular patches). Recent interest in asymmetric patches has been caused by the prediction and observation of the chiroptical effect for such patches,15 which provides new applications for particles with patches of asymmetric shapes. Many fabrication methods for patchy particles have been reported.1,9,16,17 However, only a few of them allow control of patch shape or give access to a variety of patch shapes. Nonspherical Au patterns of lower symmetry have been constructed by Zhang et al.18 on colloidal particles using particle layers with various packing patterns as masks. Patch shapes accessible by this method are limited by the available mask patterns, and the patch uniformity depends strongly on the © 2013 American Chemical Society

Received: July 4, 2013 Published: December 6, 2013 15755

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761

Langmuir

Article

part because of the lack of fabrication methods that accurately control patch shape. In the meantime, researchers have also used theoretical and simulation approaches to predict assembled structures and to study the factors that affect the assembly of patchy particles.24−30 Only recently, simulation results by Sciortino et al.25 revealed that the assembly of patchy particles is strongly affected by subtle changes in patch orientation and shape, which emphasizes the importance of exquisite control over patch shape and the need for asymmetric or less-symmetric patch shapes. In their calculations, Sciortino et al.25 showed that staggered triangular patches on opposite sides of a particle will lead to the selection of crystal structures with a staggered geometry, whereas patches arranged in an eclipsed geometry can form crystal structures only with eclipsed geometry. This simulation work and the reported chiroptical effect encouraged us to study the accessibility of triangular patch shapes via the template-assisted GLAD method. In a previous communication,31 we introduced a template for the GLAD method enabling (i) the fabrication of patchy particles with uniform hemispherical patches, (ii) the reuse of the template, and (iii) the scalability of the GLAD method. In this work, we investigate the effects of template rotation on patch size and shape, leading to well-defined but asymmetric patches and the inclusion of a second evaporation leading to overlapping patches. A straightforward geometric model predicts and confirms the patch shape, size, and position and is validated by experiments. The patch shape is determined by the glancing angle, the shadowing from neighboring particles, and the template groove resulting in beltlike and triangular patches, whereas overlapping beltlike patches are produced by consecutive evaporation and may enable the fabrication of circular patches that span the equator of the colloidal particle. In addition, preliminary magnetic field assembly results are included for particles with asymmetric triangular patches to demonstrate our ability to resuspend the patchy particles from the template after modification. Overall, we hope that our work will provide input for computational simulation and photonics research and stimulate interesting new structural predictions for colloidal assemblies and the investigation of optical properties of patchy particles.

■

Figure 1. Schematic of metal vapor flow in template-assisted patchy particle fabrication using the glancing angle deposition method with a template at various angles. (A) Single evaporation for the formation of a single patch: I0 shows the metal vapor flow at an angle of incidence θ perpendicular to the grooves; I1 shows the metal vapor flow at the same θ after the rotation of the template by angle α. (B) Two subsequent evaporations for the formation of multiple patches: I1 and I2 represent the directions of the two incident metal vapor flows at angles of incidence θ1 and θ2 after the rotation of the template by angles α1 and α2, respectively.

denoted by subscripts 1 and 2, and the rotational angle is given with respect to the projection of I0. After evaporation, the patch shapes and areas are characterized using a scanning electron microscope (EVO40 Zeiss) in high-resolution mode. For particle removal, the grooved substrate with the patchy particles is placed upside down into a cleaned glass Petri dish. One milliliter of deionized (DI) water is added to the Petri dish and exposed to sonication for 30−60 s. Subsequently, the suspension is stored in a glass vial. Prior to magnetic field assembly tests, the particle solution is concentrated by centrifugation at 3500 rpm for 5 min followed by removal of most of the supernatant. The concentrate particle suspension (40 μL) is added to a cell, covered with a microscope glass slide, allowed to equilibrate for 5 min, and exposed to the U-shaped magnet as described in earlier work.32 The patch size (coated area percentage, A%) on the PS particles is measured using the following approach. The surface of a sphere is divided evenly by a set of planes that pass through the center of the sphere, where the angle between each two adjacent planes is 360°/m and m is an integer. Furthermore, the sphere surface can also be divided evenly by a set of planes that are perpendicular to the cross section of the sphere with the distance between each two adjacent planes being r/n (the first plane goes through one of the poles of the sphere), where r is the radius of the sphere and n is an integer. Sectioning a sphere in this fashion leads to a surface that is divided into 2 × m × n sections with an identical surface area. Now, one point is chosen from each section, and its orthogonal projection onto a plane perpendicular to the cross section is obtained such that each of these orthogonal projection points corresponds to one piece of the sphere surface transferring the calculation of the sphere surface area to counting the number of points on a plane. For the SEM images of patchy particles, the radius of the particle is first determined in pixels. On the basis of the value of the radius, the coordinates of each orthogonal projection point are calculated. The area of the patch is then determined by counting the number of orthogonal projection points covered by the patch. The precision of the patch surface area is improved by increasing m and n. See Supporting Information for more details.

EXPERIMENTAL DETAILS

A template with a V-shaped groove (width = 3.7 ± 0.1 μm, groove pitch = 2.8 ± 0.1 μm) as reported in our previous work is used.31 Sulfated polystyrene particles (Molecular Probes, 1.500 ± 0.026 μm, 8.0% w/v) are loaded into the template using a convective assembly method that pulls an angled glass slide over the template with the grooves oriented perpendicular to the direction of pulling. The gold or iron patch of 50 ± 5 nm thickness is deposited on the surface of the PS particles using a benchtop evaporation system (Cressington 308R, Ted Pella, Inc.) at a pressure of 10−6 mbar.20 In the case of iron oxide patches, the iron oxide is evaporated in the presence of a 3:1 Ar/O2 mixture at 10−3 mbar.32 The direction of the incident metal vapor is varied from the perpendicular source-groove geometry previously used by rotation of the template within the plane of the sample stage. Figure 1 shows a schematic representation of particles in a V-shaped groove indicating the angle of incidence of the metal vapor flow, θ, defined as the angle between the metal vapor flow and the Si wafer surface and the rotational angle, α, measured as the angle between the orthogonal projections (white lines) of the specific evaporation direction (I1, I2) and the perpendicular direction (I0) for a single evaporation (Figure 1A) and two subsequent evaporations (Figure 1B). For the perpendicular direction, I0, α = 0°. In Figure 1B, I1 represents the direction of the first metal vapor deposition, whereas I2 represents the direction of the second metal vapor deposition. The angles of incidence for the two evaporations are

■

RESULTS AND DISCUSSION In this section, we first present our results from the geometric model and discuss how the patch shape is affected by the various parameters that are accessible in template-assisted GLAD with rotation. Next, we present experimental validation of the geometric model and preliminary magnetic field assembly results confirming that particles can be made on a large scale and can be resuspended after modification without difficulty. Last, we 15756

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761

Langmuir

Article

shadowing point and the groove point functions, the shadowing effects of both the groove and the adjacent particles cut off part of the hemispherical cap, resulting in asymmetric patches. Patches 2−4 are obtained at θP2 = 5° and αP2 = 70°, θP3 = 25° and αP3 = 50°, and θP4 = 40° and αP4 = 50°, respectively. Generally, small θ and large α give smaller patches (patch 2), whereas large θ and small α result in larger patches (patch 4). At points beyond the groove point function, the patch shape is affected only by shadowing from the adjacent particle, resulting in a symmetric crescent-shaped patch. Patches 5 and 6 are obtained at θP5 = 5° and αP5 = 88° and θP6 = 40° and αP6 = 85°, respectively. At smaller θ, a thinner patch shape (patch 5) is observed than at larger θ (patch 6). As illustrated previously,31 at θ > 54.7° a patch is formed on the particle surface, even at α = 0°. When θ > 28.7°, the height of the hemispherical patch will be larger than the radius of the particles, which means that a very small α will result in a shadow effect from the adjacent particle and affect the patch shape, resulting in an overlap of the patching and shadowing point functions at θ > 28.7°. This specific rotational angle, denoted as θC2, is affected by the groove width, w, and particle diameter, d, and can be calculated with eq 1:

discuss the advantages and limitations of the template-assisted GLAD with rotation method and potential applications. Modeling Results. Using a straightforward geometric model, we report the types of asymmetric shapes accessible via the template-assisted GLAD method with rotation. We show that the patch shape is affected by the angle of incidence, the rotational angle, the particle size, and the template dimensions and explain under which conditions an asymmetric or symmetric patch is formed. For a specific angle of incidence θ, as we increase the rotational angle from α = 0° (Figure 1A, I0) to α = 90°, where the projection of I1 is parallel to the grooves, three characteristic points are transitioned: the point at which α is large enough for the metal vapor to deposit a patch on the PS particle surface (patching point), the point at which α is large enough that the adjacent PS particle affects the patch shape (shadowing point), and the point at which α is large enough that the patch shape is not affected by the templating groove (groove point). Figure 2 plots the theoretical patching, shadowing, and groove points for 2.0-μm-diameter PS particles in a V-shaped groove of

θC2 = 54.7° − tan−1

w cos 54.7 °

d − d tan 54.7°

(1)

Owing to the fact that the template can be reused, various particle sizes can be modified using the same template. The size of the particle affects the patching point and shadowing point functions. Figure 3 depicts the trends predicted for the two

Figure 2. Theoretically predicted patch shapes obtainable by single evaporation as a function of rotational angle, α, and angle of incidence, θ, for particles with a diameter of 2.0 μm in a V-shaped groove of 4.0 μm width. Empty red triangles indicate the conditions under which a patch begins to form on the particle surface (patching point). Solid red triangles show the conditions under which neighboring particles begin to affect the patch shape (shadowing point). The solid black squares indicate α and θ for which the patch shape is no longer affected by the templating groove (groove point). Images show modeled patch shapes (blue lines) obtained at various angles of incidence and rotational angle combinations: patch 1 at θ = 25° and α = 6°, patch 2 at θ = 5° and α = 70°, patch 3 at θ = 25° and α = 50°, patch 4 at θ = 40° and α = 50°, patch 5 at θ = 5° and α = 88°, and patch 6 at θ = 40° and α = 85°.

Figure 3. Theoretically predicted patch shape diagrams for single evaporation as a function of particle diameter in a V-shaped groove of 4.0 μm width. Particle diameters are 0.8 (circles), 1.4 (diamonds), and 2.0 μm (triangles). Open symbols represent the patching point, and solid symbols indicate the shadowing point, whereas the solid black squares show the groove point.

4.0 μm width as functions of angles of incidence and rotational angles. The patching (empty red triangles), shadowing (solid red triangles), and groove point functions (solid black squares) divide the plot area into four distinct regions. Typical patch shapes predicted for the four regions are shown schematically as the area defined by the blue lines on spherical particles when viewing the patch from the metal vapor source. At θ and α in the region below the patching point function, no patch is formed on the particle surface. Between the patching point and the shadowing point functions, symmetric patches with the typical hemispherical cap caused by shadowing from the groove are formed.31 Larger θ and α angles lead to a larger patch. Patch 1 is obtained at θP1 = 25° and αP1 = 6°. In the region between the

functions with open and solid symbols, respectively, when the particle diameter is decreased from 2.0 (red triangles) to 1.4 (green diamonds) to 0.8 μm (blue circles). The groove point function (solid black squares) is not affected by the particle size because it is determined only by the groove geometry. From Figure 3 we can conclude that a reduced particle size requires a larger α at a constant θ for the preparation of patchy particles, whereas at a constant α we need a larger θ to obtain patchy particles with decreasing particle size. Furthermore, the patching point and shadowing point functions nearly overlap (the 15757

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761

Langmuir

Article

maximum coincides with the groove point (dashed line in Figure 4). Although we cannot prove that these two points are indeed the same as a result of the limited degree of precision attainable with our calculations, a qualitative explanation can be given for their coinciding. As α increases, the height of the hemispherically shaped patch increases, resulting in an increased patch size. At the same time, the particle portion shadowed by the adjacent particle is also increasing, thereby decreasing the patch size. The rate at which the patch size increases as a result of the groove’s shadowing effect is larger than the rate at which the patch size decreases as a result of shadowing from the adjacent particle. As a result, the largest patch size occurs at the groove point because beyond this point the patch is no longer affected by the groove, which also explains the decreased rate of change in A% beyond the shadowing point. Overall, the variation in A% as a function of α is small, ranging from 5.8% for θ = 45° to 12.2% for θ = 15°. The implication of this finding is that the angle of incidence, θ, sets a basic A% for the patch size that can then be varied within a 9.0 ± 3.2% range by changing the rotational angle, α. The solid line in Figure 4 indicates the shadowing points for various θ values. The patch is highly symmetric at angles below this line (Figure 2, patch 1), whereas the observed change in slope is a result of the shadowing effect of the adjacent particle, rendering the patch shape asymmetric for points beyond this line (Figure 2, patches 2−4). Figure 4B presents A% at three angles of incidence for a 1.6 μm particle in a groove of 4.0 μm width. It is included here to show that smaller particle sizes show very similar trends but show an increased A% range accessible (e.g., 11.8% vs 24.2% at θ = 20°) in the same range of rotational angles. It also shows that patch formation at low θ occurs at higher rotational angles for smaller particles. Experimental Validation and Feasibility. By changing the rotational angle and angle of incidence, various patches have been fabricated on 1.500 ± 0.026 μm polystyrene particles. In addition, subsequent metal vapor deposition has been used to add a second patch of a different material (iron), which partially overlaps the first patch (gold). Figure 5 shows patchy particles fabricated at different angle of incidence, θ, and rotational angle, α, combinations for a single gold evaporation (Figure 5A−C) and for two consecutive evaporations of gold and iron (Figure 5 D−F). It also shows the corresponding top views (row ii) for all six cases and the side views (row iii) for the single evaporation case along the vapor deposition direction obtained from the computational model. It is apparent that the patch shape and size are strong functions of both θ and α. The patches in Figure 5A(i) are fabricated at θ = 10° and α = 68°, whereas the patches in Figure 5B(i),C(i) are made at θ = 15° and α = 68° and θ = 15° and α = 78°, respectively. The SEM images enable the identification of the sharp leading and diffuse lagging boundaries,19 which result from the shadowing effect of adjacent particles and the groove, respectively. The A% values of the experimental patches shown in Figure 5A−C (row i) are 6.4, 15.7, and 16.6%, respectively. The experiments are in good agreement with the conclusions drawn from Figures 2 and 3: the patch size increases at larger angles of incidence (θ = 10 vs 15°) when the rotational angle is fixed (α = 68°), and when α is increased, the patch size is affected more by the adjacent particle and less by the groove. The calculated patch sizes for the patches shown in Figure 5 (rows ii and iii) are 7.7, 14.7, and 21.6%, in good agreement with the experimental patch areas. The somewhat larger deviation for the experimental and calculated patches shown in Figure 5C is the

rotational angle difference between these two points is less than 3°) at large α below a characteristic θC1, which increases with decreasing particle size. For particles with diameters of d = 2.0, 1.4, and 0.8 μm, the characteristic angles of incidence θC1 are 1.2, 12.6, and 26.5°, respectively, in a V-shaped groove of 4 μm width. The overlap of these two functions implies that it is difficult to make patches with the patch 1 shape (Figure 2), which is affected only by the groove. Furthermore, above a second characteristic angle of incidence, θC2, and small α, the patching point and shadowing point functions overlap again up to θ = 54.7° dictated by the groove geometry. θC2 increases from 28.7 to 38.9 to 46.9° with decreasing radius. The range of angles of incidence between θC1 and θC2 and the area beneath the groove point function and above the shadowing point function suggest that for better control of the patch shape and size and a wide range of asymmetric patch shapes a larger particle is preferred in practice. The maximum particle size for which the patch shape and size is controlled entirely by the groove is given by the constraint that the particle has to be flush with the template surface. For grooves of 4.0 μm width, the maximum particle diameter is 2.06 μm (51.8% of the groove width). The patch shape and size (i.e., the coated area percentage, A%) are a function of the angle of incidence and rotational angle. Figure 4 shows the calculated A% values as a function of α at varying θ for particles with diameters of 2.0 μm (Figure 4A) and 1.6 μm (Figure 4B) in V-shaped grooves of 4.0 μm width. θ is changed from 5 to 50° in 5° steps as indicated by the change in color and symbols used for plotting. All functions have in common that with increasing α, A% first increases up to a maximum and then decreases. It is interesting that the patch

Figure 4. Calculated coated-area percentage, A%, plotted as a function of rotational angle, α, at different angles of incidence, θ, for particles of (A) 2.0 and (B) 1.6 μm diameter in a V-shaped groove of 4.0 μm width. The solid black line shows the shadowing point, whereas the dashed line indicates the groove point. 15758

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761

Langmuir

Article

Figure 5. Scanning electron microscope images of 1.500 ± 0.026 μm single-patch polystyrene patchy particles in V-shaped grooves of 3.7 μm width after (A−C) single gold evaporation and (D−F) two evaporations of gold and iron at (i) varying angles of incidence and rotational angles and (ii) top and (iii) side views of modeled patch shapes. Patches are fabricated at (A) θ = 10° and α = 68°; (B) θ = 15° and α = 68°; (C) θ = 15° and α = 78°; (D) θ1 = 15° and α1 = 78°, θ2 = 10° and α2 = 250°; (E) θ1 = 15° and α1 = 78°, θ2 = 15° and α2 = 250°; and (F) θ1 = 20° and α1 = 45°, θ2 = 14° and α2 = 240°.

Figure 6. Magnetic field assembly of patchy particles with iron oxide patches shown in Figure 5B. (A) SEM image of patchy particles in the template after iron oxide evaporation at θ = 15° and α = 68°. The scale bar is 2 μm. (Inset) Enlarged version of a single patchy particle. The scale bar is 0.5 μm. (B) Optical microscope image of patchy particles assembled in a magnetic field of 0.008 T. The scale bar is 40 μm. (C) Close-ups of six typical chains observed during magnetic field assembly. (D) SEM image of an assembled chain after drying of the magnetic field assembly. The scale bar is 1 μm.

to assemble the particles into chains (see below). In addition, gold has a greater electron density and therefore shows a brighter contrast in SEM. The patches in Figure 5D are fabricated at θ1 = 15°, α1 = 78° and θ2 = 10°, α2 = 250°, whereas those in Figure 5E,F are prepared at θ1 = 15°, α1 = 78°, θ2 = 15°, α2 = 250° and θ1 = 20°, α1 = 40°, θ1 = 14°, α1 = 240°, respectively. In the SEM images, the brighter patch (red line) is the first patch made of gold, and the dimmer one is the second patch (blue line) made from iron. Note that because of the fact that these patches are very thin (50 ± 5 nm) and the SEM probes deeply enough to

result of a portion of the patch being out of the line of sight for the SEM as shown in the side view in Figure 5C(iii). Figure 5D−F shows a set of two-patch patchy particles with overlapping patches obtained by two sequential evaporations and their corresponding predicted patch shape. The first evaporation employed gold, whereas in the second evaporation iron was used. These two metals were chosen because they have very distinct properties. Gold is easily ligated with thiols and thus can provide anchor points for additional orthogonal assembly, whereas iron and its oxides can easily be magnetized and therefore can be used 15759

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761

Langmuir

■

Article

CONCLUSIONS By rotating the template to specific angles in conjunction with changing angle of incidences, patchy particles with an abundance of asymmetric patch shapes, triangular patches as simulated by Sciortino et al.,25 and overlapping patches have been successfully fabricated. Even beltlike and crescent moon-shaped patches can be achieved, which enriches the parameter space for patchy particle fabrication and assembly and the study of their properties. Initial assembly tests validate particle removal from the template and show an unprecedented branching of chains. Template-assisted GLAD with the rotational method also enables the fabrication of patchy particles with overlapping patches, in which the amount of overlap can be varied by changing θ and α and can be predicted using the computational model. As a new feature, template-assisted GLAD with rotation enables asymmetric patch shapes and the partial modification of the lower particle surface.

sample both patches in the area where they overlap, the gold patch dominates the image. We have added our predicted patch shapes to one particle in each image to help guide the eye. Row ii in Figure 5D−F depicts the predicted patch shapes when the particle is viewed from the top of the template. The experimental A% values of the patches shown in row i are determined to be 23.8, 27.9, and 18.2%, respectively. The corresponding calculated A% values for the predicted patches shown in row ii are 28.6, 33.4, and 17.5%, respectively. The deviation between the experimental and computational patch sizes in Figure 5D,E is again due to the hidden portion of the gold patch (red line). Furthermore, from the computed patch shapes, the overlapping parts of the two patches are calculated to be 4.4, 7.3, and 6.8%. From a comparison of Figure 5D,E, we can determine that increasing θ2 leads to more overlap and an overall increase in A%. Figure 6 summarizes the first attempts at assembling patchy particles with identical iron oxide patches of the type shown in Figure 5B in a magnetic field of 0.008 T. Iron is deposited at a rate of 0.35 nm/s in a 3:1 Ar/O2 mixture to a thickness of 79.2 nm, yielding Fe3O4 patches32 with a triangular shape (Figure 6A). Upon application of the magnetic field, the particles quickly assemble into chains along the magnetic field lines (Figure 6B). In contrast to the double chains observed for Fe3O4 Janus particles,32 the patchy particle suspension yields chains that are branched and not close-packed (Figure 6C). Drying of the assembled chains allows imaging of the patch orientation within a chain (Figure 6D). The connectivity of the patches is apparent from the SEM image; however, the orientation of the patches appears to be random along the chain with a tendency to align the patch such that its longest axis is in the direction of the chain. Further experimental and theoretical studies are underway to understand the driving forces leading to these assembled structures. We have shown that template-assisted GLAD gives access to a large variety of asymmetric patches with controlled patch shapes when the rotation of the template is enabled and θ and α are chosen such that the patch is produced between the shadowing and groove points. Besides the asymmetric, triangular patch shapes, hemispherical and crescent moon-shaped patches are also accessible. The abundance of shapes provides more choices for studying the effect of patch shape on patch interactions and assembly properties. Assembly tests prove that patchy particles can be easily resuspended from the template after modification through a short exposure to sonication and can be produced in large enough quantities for assembly studies. In addition, preliminary magnetic field assembly of Fe3O4 patchy particles indicates that asymmetric patch shape leads to the ability to form branched chains. In the case of two-patch patchy particles with overlapping patches, the overlapping part can be varied by changing the angle of incidence and rotational angle and is predicted using the computational model. Such particles with overlapping patches will be important in assembly with orthogonal fields and the production of chiral patches, which are topics of current investigations in our laboratory. Promoted by the ability to fabricate patches on the surface that cannot be seen from the top, template-assisted GLAD with rotation enables the modification of the lower particle surface when the particle cannot be turned easily. One of the more interesting shapes that can be obtained with this approach is two long beltlike patches that partially overlap and create a patch with a partial circular shape around the waist of the particle with the two poles staying unmodified. Such patches have promising applications in the field of autonomous motion.12,33

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information on patch area calculations. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This project was supported by the National Science Foundation under CAREER award NSF-CBET 0644789. REFERENCES

(1) Charnay, C.; Lee, A.; Man, S.-Q.; Moran, C. E.; Radloff, C.; Bradley, R. K.; Halas, N. J. Reduced Symmetry Metallodielectric Nanoparticles: Chemical Synthesis and Plasmonic Properties. J. Phys. Chem. B 2003, 107, 7327−7333. (2) Ramírez, L. M.; Milner, S. T.; Snyder, C. E.; Colby, R. H.; Velegol, D. Controlled Flats on Spherical Polymer Colloids. Langmuir 2009, 26, 7644−7649. (3) Yake, A. M.; Snyder, C. E.; Velegol, D. Site-Specific Functionalization on Individual Colloids: Size Control, Stability, and Multilayers. Langmuir 2007, 23, 9069−9075. (4) Chaudhary, K.; Chen, Q.; Juárez, J. J.; Granick, S.; Lewis, J. A. Janus Colloidal Matchsticks. J. Am. Chem. Soc. 2012, 134, 12901−12903. (5) Chastek, T. T.; Hudson, S. D.; Hackley, V. A. Preparation and Characterization of Patchy Particles. Langmuir 2008, 24, 13897−13903. (6) Bao, H.; Butz, B.; Zhou, Z.; Spiecker, E.; Hartmann, M.; Klupp Taylor, R. N. Silver-Assisted Colloidal Synthesis of Stable, Plasmon Resonant Gold Patches on Silica Nanospheres. Langmuir 2012, 28, 8971−8978. (7) McConnell, M. D.; Kraeutler, M. J.; Yang, S.; Composto, R. J. Patchy and Multiregion Janus Particles with Tunable Optical Properties. Nano Lett. 2010, 10, 603−609. (8) Chen, Q.; Diesel, E.; Whitmer, J. K.; Bae, S. C.; Luijten, E.; Granick, S. Triblock Colloids for Directed Self-Assembly. J. Am. Chem. Soc. 2011, 133, 7725−7727. (9) Chen, Q.; Yan, J.; Zhang, J.; Bae, S. C.; Granick, S. Janus and Multiblock Colloidal Particles. Langmuir 2012, 28, 13555−13561. (10) Chen, Q.; Bae, S. C.; Granick, S. Directed Self-Assembly of a Colloidal Kagome Lattice. Nature 2011, 469, 381−384. (11) De Michele, C.; Bellini, T.; Sciortino, F. Self-Assembly of Bifunctional Patchy Particles with Anisotropic Shape into Polymers

15760

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761

Langmuir

Article

Chains: Theory, Simulations, and Experiments. Macromolecules 2011, 45, 1090−1106. (12) McDermott, J. J.; Velegol, D. Simple Fabrication of Metallic Colloidal Doublets Having Electrical Connectivity. Langmuir 2008, 24, 4335−4339. (13) Ramírez, L. M.; Smith, A. S.; Unal, D. B.; Colby, R. H.; Velegol, D. Self-Assembly of Doublets from Flattened Polymer Colloids. Langmuir 2012, 28 (), 4086−4094. (14) Wang, Y.; Wang, Y.; Breed, D. R.; Manoharan, V. N.; Feng, L.; Hollingsworth, A. D.; Weck, M.; Pine, D. J. Colloids with Valence and Specific Directional Bonding. Nature 2012, 491, 51−55. (15) Hou, Y.; Li, S.; Su, Y.; Huang, X.; Liu, Y.; Huang, L.; Yu, Y.; Gao, F.; Zhang, Z.; Du, J. Design and Fabrication of Three-Dimensional Chiral Nanostructures Based on Stepwise Glancing Angle Deposition Technology. Langmuir 2012, 29, 867−872. (16) Jiang, S.; Granick, S. A Simple Method to Produce Trivalent Colloidal Particles. Langmuir 2009, 25, 8915−8918. (17) Lin, C.-C.; Liao, C.-W.; Chao, Y.-C.; Kuo, C. Fabrication and Characterization of Asymmetric Janus and Ternary Particles. ACS Appl. Mater. Interfaces 2010, 2, 3185−3191. (18) Zhang, G.; Wang, D.; Möhwald, H. Patterning Microsphere Surfaces by Templating Colloidal Crystals. Nano Lett. 2004, 5 (), 143− 146. (19) Pawar, A. B.; Kretzschmar, I. Patchy Particles by Glancing Angle Deposition. Langmuir 2007, 24, 355−358. (20) Yu, Y.; Gan, L.; Zhang, G.; Yang, B. Asymmetric Microparticles and Heterogeneous Microshells via Angled Colloidal Lithography. Colloids Surf., A 2012, 405 (), 51−58. (21) Imura, Y.; Kato, M.; Kondo, T.; Kawai, T. Strings of Metal HalfShells Fabricated Using Colloidal Particle Monolayer as a Template. Langmuir 2010, 26 (13), 11314−11318. (22) Halder, A.; Kundu, P.; Viswanath, B.; Ravishankar, N. Symmetry and Shape Issues in Nanostructure Growth. J. Mater. Chem. 2010, 20, 4763−4772. (23) Shah, A. A.; Schultz, B.; Kohlstedt, K. L.; Glotzer, S. C.; Solomon, M. J. Synthesis, Assembly, and Image Analysis of Spheroidal Patchy Particles. Langmuir 2013, 29, 4688−4696. (24) Romano, F.; Sciortino, F. Colloidal Self-Assembly: Patchy from the Bottom Up. Nat. Mater. 2011, 10, 171−173. (25) Romano, F.; Sciortino, F. Patterning Symmetry in the Rational Design of Colloidal Crystals. Nat. Commun. 2012, 3, 975. (26) Zhang; Glotzer, S. C. Self-Assembly of Patchy Particles. Nano Lett. 2004, 4, 1407−1413. (27) Pawar, A. B.; Kretzschmar, I.; Aranovich, G.; Donohue, M. D. SelfAssembly of T-Structures in Molecular Fluids. J. Phys. Chem. B 2007, 111 (), 2081−2089. (28) Giacometti, A.; Lado, F.; Largo, J.; Pastore, G.; Sciortino, F. Effects of Patch Size and Number within a Simple Model of Patchy Colloids. J. Chem. Phys. 2010, 132, 174110. (29) Doppelbauer, G.; Bianchi, E.; Kahl, G. Self-Assembly Scenarios of Patchy Colloidal Particles in Two Dimensions. J. Phys.: Condens. Matter 2010, 22, 104105. (30) Sajanlal, P. R.; Sreeprasad, T. S.; Samal, A. K.; Pradeep, T. Anisotropic Nanomaterials: Structure, Growth, Assembly, and Functions. Nano. Rev. 2011, 2, 1−62. (31) He, Z.; Kretzschmar, I. Template-Assisted Fabrication of Patchy Particles with Uniform Patches. Langmuir 2012, 28, 9915−9919. (32) Ren, B.; Ruditskiy, A.; Song, J. H.; Kretzschmar, I. Assembly Behavior of Iron Oxide-Capped Janus Particles in a Magnetic Field. Langmuir 2011, 28, 1149−1156. (33) Pawar, A. B.; Kretzschmar, I. Multifunctional Patchy Particles by Glancing Angle Deposition. Langmuir 2009, 25, 9057−9063.

15761

dx.doi.org/10.1021/la404592z | Langmuir 2013, 29, 15755−15761