Registered charity number: 207890

Highlighting the joint research results from the Soft Matter and

Nano Engineering Lab of Dr Shan Jiang at Iowa State University,

Dr Stephen M. Anthony at Sandia National Laboratories and

the Soft Matter and Interfacial Phenomena Lab of Dr Xin Yong

at Binghamton University.

Drying mediated orientation and assembly structure of

amphiphilic Janus particles

Amphiphilic Janus particles demonstrate unique assembly

structures when dried on a substrate. Miller et al. detail

simulation-backed experimental results as the first step in

understanding the drying process involved with Janus particles.

The study lays out the foundation of using Janus particles as

coating materials to effectively change the surface properties.

rsc.li/soft-matter-journal

As featured in:

See Xin Yong, Shan Jiang et al.,Soft Matter, 2018, 14, 6793.

This journal is©The Royal Society of Chemistry 2018 Soft Matter, 2018, 14, 6793--6798 | 6793

Cite this: SoftMatter, 2018,

14, 6793

Drying mediated orientation and assemblystructure of amphiphilic Janus particles†

Kyle Miller, a Ayuna Tsyrenova,a Stephen M. Anthony, b Shiyi Qin,c

Xin Yong *c and Shan Jiang *ad

Amphiphilic Janus particles demonstrate unique assembly struc-

tures when dried on a hydrophilic substrate. Particle orientations

are influenced by amphiphilicity and Janus balance. A three-stage

model is developed to describe the process. Simulation further

indicates the dominant force is capillary attraction due to the

interface pinning at rough Janus boundaries.

The drying of colloidal suspensions is a ubiquitous process innature and everyday life. Many commercial applications,including printing and coating products,1,2 rely on drying toform a coherent thin film that adheres to the substrate.Furthermore, the ability to fabricate functional thin films viaa simple drying process will help address challenges in cutting-edge applications such as flexible electronics, bio-sensing, andmultifunctional coatings.3–5 From a fundamental perspective,the drying process usually involves fast dynamics that are farfrom equilibrium. In addition, the liquid–gas interface furthercomplicates the transport, interaction, and assembly of thecolloidal constituents during the process. Previous studies haverevealed interesting patterns driven by evaporation. The primeexample is the ‘‘coffee ring’’ commonly observed in dryingsessile droplets.6 However, most of the studies have onlyfocused on homogeneous or isotropic particles and their spatialdistribution. No systematic studies, in either experiment orsimulation, have reported how structured particles, such asJanus particles, where two sides of the particles are made of

different chemical compositions, self-assemble during the dryingprocess.

More importantly, the anisotropy of Janus particles elicits acompletely new question regarding particle orientation, whichdoes not exist for conventional isotropic particles. Orientationis also associated with anisotropic molecules such as surfactants,proteins, and di-block copolymers. Studies have shown thatdrying conditions can drastically affect the assembly andorientation of these molecules and even induce different phasebehaviors.7–9 Beyond the molecular systems, Janus nano-particles and nanoclusters demonstrate unique assembly struc-tures both in bulk and at interfaces.10,11 However, the assemblyand orientation of Janus particles at much larger length scales(mm) under drying conditions have not been extensively studied.Due to the anisotropy of interactions and amphiphilicity ofJanus particles, their interactions among each other and withinterfaces are completely different from homogeneous particles.The interface and the drying process will play a significant rolein controlling the assembly and orientation of Janus particles.This report offers a new model system in both experiment andsimulation for providing insight into the unique behavior ofanisotropic building blocks. Furthermore, the orientation ofJanus particles is a critical consideration for their functionality,as it directly affects the material properties of the dried film.4

In this report, we examined the translational and rotationaldynamics of Janus particles during the drying process. Particleorientation was analyzed using an image analysis algorithmon scanning electron microscope (SEM) images, which showcontrast between the coated sides. We propose a three-stagemodel to describe the drying process of Janus particles. Wediscovered that the orientations and assembly structures ofJanus particles are affected by both amphiphilicity and Janusbalance (geometry) of spheres. Additionally, particle orientationis highly coupled with aggregation structures. A coarse-grainedsimulation with explicit solvent was developed to reveal thefundamental interactions between particles during drying.Capillary attraction was induced by pinning the contact linearound the rough boundaries on Janus particles. This force is

a Department of Materials Science and Engineering, Iowa State University of Science

and Technology, Ames, IA 50011, USA. E-mail: sjiang1@iastate.edub Department of Bioenergy and Defense Technology,

Sandia National Laboratories, Albuquerque, NM 87123, USAc Department of Mechanical Engineering, Institute for Materials Research,

Binghamton University, Binghamton, NY 13902, USA.

E-mail: xyong@binghamton.edud Division of Materials Science & Engineering, Ames National Laboratory, Ames,

IA 50011, USA

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8sm01147h

Received 4th June 2018,Accepted 27th June 2018

DOI: 10.1039/c8sm01147h

rsc.li/soft-matter-journal

Soft Matter

COMMUNICATION

6794 | Soft Matter, 2018, 14, 6793--6798 This journal is©The Royal Society of Chemistry 2018

found to dominate the interactions and leads to unique deposi-tion patterns.

Janus particles were fabricated following a protocol developedpreviously.12 Experimental details can be found in ESI.† Au (toplayer) was deposited onto a monolayer of 3 mm silica particles.Thiol molecules are then used to render the Au surface withdifferent chemical properties. ODT (1-octadecanethiol) and DDT(1-dodecanethiol) are used to render the Au surface hydrophobicand obtain amphiphilic Janus particles (JP-ODT and JP-DDT),while MHA (16-mercaptohexadecanoic acid) is used to give theAu surface a negative charge (JP-MHA). To obtain Janus particlesof different Janus balance, coated Janus particles were etchedby aqua regia solution. Control particles (JP-SiO2), which areoptically Janus however chemically homogeneous, were obtainedby depositing another layer of silica onto the Au coated particlemonolayer instead of the thiol treatment.

For the drying experiment, we only used hydrophilic glasssubstrate cleaned by O2 plasma. The drying process was imagedin real time via an optical microscope. The details of driedparticle assembly were imaged using SEM. Both backscatteredand secondary electron images were taken and fed intoan algorithm (ESI†) developed to extract orientation of Janusparticles.

For computer simulations of evaporating colloidal films, weperformed many-body dissipative particle dynamics (MDPD)13–16

to probe competing interactions that drive the self-assembly ofJanus particles. MDPD models solvent and Janus particlesexplicitly by representing a volume of fluid or solid as individualcoarse-grained beads, whose dynamics are governed by classicalmechanics. Similar to standard dissipative particle dynamics(DPD),17–19 MDPD captures hydrodynamics in colloidal suspen-sions. However, different from DPD, which can model onlysingle-phase fluid or density-matching binary fluid, MDPD iscapable of modeling multiphase fluid with large density andviscosity contrasts, e.g., liquid–vapor coexistence (Fig. 3). Thus,the drying process can be explicitly simulated using MDPD.Evaporation is modeled by reducing the vapor density belowthe saturation vapor density through continuous removal of vaporbeads.20 The details of the evaporation model are described inESI.† Colloidal particles are modeled as non-deformable clustersof frozen MDPD beads translating and rotating as rigidbodies.21–23 The particle bead density is adjusted to prevent thepenetration between solvent beads and particles. In addition,since adsorption energy (106 kT) is magnitudes higher than thegravity energy (102 kT) associated with the center of weight shiftdue to the gold coating in the experiment, the density of gold isnot considered in the simulation.24 The surface chemistry of eachparticle is readily controlled by tuning the interaction parametersof the constituent beads.25–27 Janus boundary roughness is intro-duced as a sinusoidal perturbation with amplitude dr as shown inFig. S1 (ESI†). Other simulation details are described in ESI.†

Drying of colloidal suspensions is a complicated process.Similar studies have focused on the lateral distribution ofisotropic particles, namely the coffee ring effect due toevaporation-induced capillary flow.6 We also observed coffeering formed by Janus particles, which is beyond the scope of

this report. Instead, we emphasize the unique orientation andlocal monolayer assembly of a dilute solution of Janus particlesmediated by the drying process.

Fig. 1 clearly shows the assembly structure and particleorientation after the solution is completely dried. The brightside is the Au coated side. For Janus particles, almost all theparticles orient with the Au side facing towards the air, without-of-plane angle (y) (Fig. S2, ESI†) very close to 01 (Fig. 1a).For homogeneous particles, obviously orientation should berandom, which is clearly demonstrated by the control particlesJP-SiO2 in Fig. 1b. Particle tracking results shown in Fig. 1c and dfurther confirmed the observation quantitatively. The orienta-tional preference for larger angles for control particles is likelydue to the high density of Au coating, which directs particles toorient more towards the substrate.

Interestingly, Janus particles form a large-scale open fractalstructure while homogeneous particles only show small aggre-gates. In separate experiments, the deposits of JP-ODT andJP-DDT showed very similar structures when dried under thesame conditions, while JP-MHA resembles the homogeneousparticles (Fig. S3, ESI†). We note that both ODT and DDT arehydrophobic ligands, while MHA is negatively charged. Theresults indicate that the fractal structures were formed due tothe amphiphilicity of Janus particles. In addition, the fractalcluster structures suggest that the assembly is limited bydiffusion and Janus particles may experience strong attractionsduring the drying process.

Real-time video revealed the detailed dynamics of the dryingprocess. The snapshots in Fig. 2 suggest the drying process hasthree stages. In Stage I, Janus particles adsorb onto theair–water interface as the water surface recedes towards thesubstrate; in Stage II, Janus particles quickly aggregate intocluster structures; in Stage III, the aggregates further assembleas water evaporates away and the assembly structures areeventually deposited onto the substrate. The particle aggregationpattern is consistent with the Janus particle assembly structuresformed at the interface. In Fig. 2, particles adsorbed at theinterface showed a much darker color, which is due to the Auside orienting completely towards the air and blocking most ofthe transmitted light compared to freely rotating particles.

Another interesting observation is that Stages I and II almosthappen simultaneously. This indicates again the strong attrac-tions between Janus particles adsorbed at the interface. Notably,both the silica side of the Janus particles and the substrate arenegatively charged, yet repulsive electrostatic interaction is notdominant. Although it is difficult to quantify, we consistentlyobserved that particles with fewer neighbors are more likely toorient randomly. This suggests that the clustering of Janusparticles is responsible for maintaining the orientation of particlesduring the Stage III of the drying process.

Previous studies have reported that the amphiphilic Janusspheres may experience strong attraction at fluid interfaces dueto the capillary force.28 Because of imperfections in fabrication,the Janus boundary possesses a certain roughness that distortsthe pinned water–air interface. The hypothesis is that whenwater pins at the rough boundary, strong attraction is induced

Communication Soft Matter

This journal is©The Royal Society of Chemistry 2018 Soft Matter, 2018, 14, 6793--6798 | 6795

between particles to minimize the total surface free energy.Since it is very hard to measure and quantify the roughness ofJanus boundaries experimentally, we carried out computersimulations to further study the particle interactions duringthe drying process. A simulation consists of an equilibriumstage to probe the self-assembly of particles at the interface inStage II and a drying stage to reveal the final deposit in Stage III.Fig. 3 shows that Janus particles adsorb strongly at the liquid–vapor interface with their orientation fixed, while the hydrophilichomogeneous particles are dispersed uniformly in the film withrandom orientation. MDPD simulations allow us to directlycharacterize the effect of Janus boundary roughness. The‘‘rough’’ Janus particles with prescribed sinusoidal boundariesassemble into a single fractal cluster at the interface, which isreminiscent of the assembly structures observed in the experi-ments. Upon complete evaporation of solvent, the large clustermaintains its fractal structure and Janus particle orientationis unchanged (Movie S1, ESI†). In contrast, despite havingthe fixed orientation, the ‘‘perfect’’ Janus particles withsmooth Janus boundaries do not exhibit long-range orderingat equilibrium. Even with the assistance of capillary bridgesbetween particles at the final stage of drying (see Movie S2,ESI†), only small aggregates can be observed in Fig. 3b. Thiscomparison highlights the critical role of the Janus boundaryroughness.

Another observation in simulation is that the homogeneousparticles also form a large cluster with random particle orienta-tion after drying (see Fig. 3c). This assembly is solely inducedby the capillary interactions in the final drying stage whenthe water film thickness drops below single particle diameter.Hydrophilic particles significantly deform the interface asshown in Movie S3 (ESI†), leading to larger attractive forces.The discrepancy between simulation and experiment is mainlydue to negligible substrate friction and smaller inter-particledistances of control particles in the simulation, which promotelarge-scale aggregation.

We quantitatively characterize the roughness-induced effectiveattractions between a particle pair (see ESI† for details). Fig. 3dconfirms the interfacial distortion induced by the rough Janusboundary and the consequent long-range capillary attractionsbetween particles. The strength of capillary force increases asthe roughness increases. In addition, the strength is sensitive tothe actual morphology of the Janus boundary, which is mani-fested by the amplitude of roughness and the particle radius inour simulations. Nevertheless, a simulation with large particlesRp = 10 with rough Janus boundaries shows similar fractalaggregates (Fig. S4, ESI†), confirming that the capillary attrac-tion leads to the unique large-scale assembly. Notably, a smallroughness (dr = 1) comparable to the interfacial thickness20

fails to induce attraction. This is because in this situation the

Fig. 1 SEM images and orientation analysis of dried particles on hydrophilic silicon wafer: (a) amphiphilic Janus particles; (b) homogeneous controlparticles, inset in the upper right corner indicates the image analysis of particle orientation; (c) orientation analysis of amphiphilic Janus particles(d) orientation analysis of homogeneous control particles.

Soft Matter Communication

6796 | Soft Matter, 2018, 14, 6793--6798 This journal is©The Royal Society of Chemistry 2018

pinning induced distortion is too small to withstand theintrinsic thermal fluctuations of the interface. We did notattempt to fit the simulation results to the well-known powerlaw F B r�5,29 since the sinusoidal roughness does not necessarilyreproduce the quadrupolar distortion of the contact line inthe experiments.30

In experiment, we further studied the drying pattern ofJanus particles with different Janus balance. The Au coatingon particles can be gradually etched away as shown in Fig. 4.

The etched Janus particles have a smaller hydrophobic patch,and the size of the patch can be fine-tuned by controlling theetching time. An angle a as defined in Fig. 4a is used tocharacterize the geometry of Janus spheres, with a = 901 forun-etched particles. A series of amphiphilic Janus particles withdifferent Janus balance were dried under the same conditions.Fig. 4c shows that Janus balance can affect the particle orienta-tion. When a o 601, particles orient more randomly, andthe assembly structures are small, closely packed aggregates.

Fig. 3 Representative simulation snapshots of the assembly structures of (a) amphiphilic particles with rough Janus boundaries, (b) amphiphilic particleswith smooth Janus boundaries, and (c) homogeneous control particles. (d) Effective interaction force between two amphiphilic Janus particles withrough boundaries. Negative magnitude corresponds to attraction. Error bars represent standard deviations in time. The inset shows the time-averagedinterfacial distortion induced by the Janus boundary roughness of adsorbed amphiphilic particles (Rp = 10 and dr = 3).

Fig. 2 Three-stage drying process: (a) schematic plot; (b) optical microscope images of different drying stages.

Communication Soft Matter

This journal is©The Royal Society of Chemistry 2018 Soft Matter, 2018, 14, 6793--6798 | 6797

The structure is very similar to homogeneous particles. However,when a4 701, most particles orient their hydrophobic side towardsair (y o 301) and form a fractal assembly. The observation can bewell explained by the capillary attraction suggested by simulation.Roughness at the boundary can still be responsible for the strongattraction for slightly etched particles (a4 701). However, when theAu coating patch becomes very small (a o 601), the water contactlines on adjacent particles are too far away from each other toinduce strong enough attraction.

In conclusion, amphiphilic Janus particles orient with thehydrophobic side facing towards air and form fractal assemblystructures when dried on a hydrophilic substrate. We furtherdiscovered that the orientation depends on the amphiphilicity andJanus balance. The orientation is maintained due to the aggrega-tion formed when Janus particles adsorb at the water–air interface.The structures suggest that Janus particles experience strongattractions when adsorbed at interface. Computer simulationsreproduced the drying process and provide convincing evidencethat the strong capillary attraction induced by the pinning of theinterface at the rough Janus boundaries dominates the system. Ifthe Janus particle boundaries are completely smooth, no strongattraction was observed in the simulation.

This research is the first step in understanding the dryingprocess involved with Janus particles. It reveals the uniquephysics associated with Janus spheres. The study also lays outthe foundation of using Janus particles as coating materials toeffectively change surface properties. The tools developed herecan serve as platform technologies for image analysis of Janusspheres and simulation of drying process.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The experiment part of this work was supported by Iowa StateUniversity Start-up Fund, and 3M Non-tenured Faculty Award.The simulation part of this work was partially supported by theAmerican Chemical Society Petroleum Research Fund undergrant no. 56884-DNI9 and used resources of the Center forFunctional Nanomaterials, which is a U.S. DOE Office of ScienceFacility, at Brookhaven National Laboratory under contract no. DE-SC0012704. This paper describes objective technical results andanalysis. Any subjective views or opinions that might be expressedin the paper do not necessarily represent the views of the U.S.Department of Energy or the United States Government. SandiaNational Laboratories is a multimission laboratory managed andoperated by National Technology & Engineering Solutions ofSandia, LLC, a wholly owned subsidiary of Honeywell InternationalInc., for the U.S. Department of Energy’s National Nuclear SecurityAdministration under contract DE-NA0003525.

References

1 B.-J. De Gans and U. S. Schubert, Inkjet Printing of Well-DefinedPolymer Dots and Arrays, Langmuir, 2004, 20(18), 7789–7793.

2 P. A. Steward, J. Hearn and M. C. Wilkinson, An Overview ofPolymer Latex Film Formation and Properties, Adv. ColloidInterface Sci., 2000, 86(3), 195–267.

3 I. I. Smalyukh, O. V. Zribi, J. C. Butler, O. D. Lavrentovichand G. C. L. Wong, Structure and Dynamics of LiquidCrystalline Pattern Formation in Drying Droplets of DNA,Phys. Rev. Lett., 2006, 96(17), 177801.

4 S. Jiang, A. Van Dyk, A. Maurice, J. Bohling, D. Fasano andS. Brownell, Design Colloidal Particle Morphology and Self-Assembly for Coating Applications, Chem. Soc. Rev., 2017,46(12), 3792–3807.

Fig. 4 Orientation and assembly affected by Janus balance: (a) Janus particles with different Janus balance synthesized by the Au etching;(b) orientation versus different Janus balance; (c) SEM images of Janus particle assembly when a = 621.

Soft Matter Communication

6798 | Soft Matter, 2018, 14, 6793--6798 This journal is©The Royal Society of Chemistry 2018

5 T. P. Bigioni, X. M. Lin, T. P. Bigioni, X. M. Lin, T. T.Nguyen, E. I. Corwin, T. A. Witten and H. M. Jaeger.,Nat. Mater., 2006, 5, 265.

6 R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nageland T. A. Witten, Capillary Flow as the Cause of Ring Stainsfrom Dried Liquid Drops, Nature, 1997, 389(6653), 827.

7 D. A. Doshi, A. Gibaud, N. Liu, D. Sturmayr, A. P. Malanoski,D. R. Dunphy, H. Chen, S. Narayanan, A. MacPhee andJ. Wang, In situ X-Ray Scattering Study of Continuous Silica –Surfactant Self-Assembly during Steady-State Dip Coating,J. Phys. Chem. B, 2003, 107(31), 7683–7688.

8 A. P. Le Brun, J. Chow, D. V. Bax, A. Nelson, A. S. Weiss andM. James, Molecular Orientation of Tropoelastin Is Deter-mined by Surface Hydrophobicity, Biomacromolecules, 2012,13(2), 379–386.

9 J. Zhang, D. Posselt, D.-M. Smilgies, J. Perlich, K. Kyriakos,S. Jaksch and C. M. Papadakis, Lamellar Diblock CopolymerThin Films during Solvent Vapor Annealing Studied byGISAXS: Different Behavior of Parallel and PerpendicularLamellae, Macromolecules, 2014, 47(16), 5711–5718.

10 Z. Liu, R. Guo, G. Xu, Z. Huang and L.-T. Yan, Entropy-Mediated Mechanical Response of the Interfacial NanoparticlePatterning, Nano Lett., 2014, 14(12), 6910–6916.

11 C. Ma, H. Wu, Z. Huang, R. Guo, M. Hu, C. Kubel, L. Yanand W. Wang, A Filled-Honeycomb-Structured Crystal Formedby Self-Assembly of a Janus Polyoxometalate-Silsesquioxane(POM-POSS) Co-Cluster, Angew. Chem., Int. Ed., 2015, 54(52),15699–15704.

12 S. Jiang, J. Yan, J. K. Whitmer, S. M. Anthony, E. Luijten andS. Granick, Orientationally Glassy Crystals of Janus Spheres,Phys. Rev. Lett., 2014, 112(21), 218301.

13 I. Pagonabarraga and D. Frenkel, Dissipative ParticleDynamics for Interacting Systems, J. Chem. Phys., 2001,115, 5015.

14 S. Y. Trofimov, E. L. F. Nies and M. A. J. Michels, Thermo-dynamic Consistency in Dissipative Particle Dynamics Simula-tions of Strongly Nonideal Liquids and Liquid Mixtures,J. Chem. Phys., 2002, 117, 9383–9394.

15 P. B. Warren, Vapor–Liquid Coexistence in Many-BodyDissipative Particle Dynamics, Phys. Rev. E: Stat., Nonlinear,Soft Matter Phys., 2003, 68, 066702.

16 M. Arienti, W. Pan, X. Li and G. Karniadakis, Many-BodyDissipative Particle Dynamics Simulation of Liquid/Vaporand Liquid/Solid Interactions, J. Chem. Phys., 2011,134(20), 204114.

17 P. J. Hoogerbrugge and J. M. V. A. Koelman, SimulatingMicroscopic Hydrodynamic Phenomena with DissipativeParticle Dynamics, EPL, 1992, 19(3), 155–160.

18 P. Espanol and P. Warren, Statistical Mechanics of Dissipa-tive Particle Dynamics, EPL, 1995, 30(4), 191–196.

19 R. D. Groot and P. B. Warren, Dissipative Particle Dynamics:Bridging the Gap between Atomistic and Mesoscopic Simulation,J. Chem. Phys., 1997, 107(11), 4423–4435.

20 X. Yong, S. Qin and T. J. Singler, Nanoparticle-MediatedEvaporation at Liquid–vapor Interfaces, Extreme Mech. Lett.,2016, 7, 90–103.

21 S. Chen, N. Phan-Thien, B. C. Khoo and X. J. Fan, Flowaround Spheres by Dissipative Particle Dynamics, Phys.Fluids, 2006, 18(10), 103605.

22 H. Fan and A. Striolo, Mechanistic Study of Droplets Coalescencein Pickering Emulsions, Soft Matter, 2012, 9533.

23 X. Yong, Modeling the Assembly of Polymer-Grafted Nanoparticlesat Oil–Water Interfaces, Langmuir, 2015, 31, 11458–11469.

24 S. Jiang and S. Granick, Janus Balance of Amphiphilic ColloidalParticles, AIP, 2007.

25 I. Salib, X. Yong, E. J. Crabb, N. M. Moellers, G. T. McFarlin,O. Kuksenok and A. C. Balazs, Harnessing Fluid-DrivenVesicles to Pick up and Drop off Janus Particles, ACS Nano,2013, 7, 1224–1238.

26 X. Yong, E. J. Crabb, N. M. Moellers and A. C. Balazs, Self-Healing Vesicles Deposit Lipid-Coated Janus Particles intoNanoscopic Trenches, Langmuir, 2013, 29, 16066–16074.

27 P. B. Warren, No–Go Theorem in Many-Body DissipativeParticle Dynamics, Phys. Rev. E: Stat., Nonlinear, Soft MatterPhys., 2013, 87(4), 045303.

28 A. Kumar, B. J. Park, F. Tu and D. Lee, Amphiphilic JanusParticles at Fluid Interfaces, Soft Matter, 2013, 9(29), 6604–6617.

29 I. B. Liu, N. Sharifi-Mood and K. J. Stebe, Capillary Assemblyof Colloids: Interactions on Planar and Curved Interfaces,Annu. Rev. Condens. Matter Phys., 2018, 9(1), 283–305.

30 B. J. Park, T. Brugarolas and D. Lee, Janus Particles at anOil–water Interface, Soft Matter, 2011, 7(14), 6413.

Communication Soft Matter