Formation of three-dimensional (3D) Self-Assembled Clusters of Anisotropic Janus Particles in Microgravity

Janus particles used in this study exhibit both chemical and geometrical anisotropy. For the chemical aspect, hydrophilicity and hydrophobicity are imparted identically across all three Janus particles. Geometrically, the ratios of hydrophobic to hydrophilic parts are set at 3:7, 5:5, and 7:3, respectively. Additionally, the convex-top geometry is uniformly applied to all three ratios. The Janus particles, with controlled hydrophobic and hydrophilic ratios, were prepared through our sequential micromolding process with slight modifications (Choi et al., 2010; Yeom et al., 2016). Unlike the conventional PDMS mold, a PFPE (Perfluoropolyether dimethacrylate (PFPE-DMA, MD-700), Solvay) mold was prepared and used for reliable particle preparation. First, a photocurable mixture composed of PFPE-DMA and 5% photoinitiator (w/w, darocur 1173 (2-hydroxy-2-methylpropiophenon, Sigma-Aldrich)) was purged by N₂ gas. This mixture was applied to a Si master with predefined SU-8 photoresist patterns (convex-top cylindrical geometry, 3 μm × 5 μm) and photopolymerized into a PFPE mold under ultraviolet (UV) irradiation for 10 minutes in a continuous N₂ gas atmosphere. The hydrophobic photocurable solution, composed of DDDMA (1,10-decanediol dimethacrylate, polyscience) and 5% (w/w) Irgacure 184 (1-Hydroxycyclohexyl phenyl ketone, Sigma-Aldrich) photoinitiator, was mixed with ethanol at three different volume ratios to prepare anisotropic Janus particles with controlled hydrophobic and hydrophilic ratios (3:7, 5:5, 7:3). The hydrophilic photocurable solution comprised 75% (v/v) PEG-DA (Poly(ethylene glycol) diacrylate, average Mn=700, Sigma-Aldrich), 20% CEA (2-Carboxyethyl acrylate, Sigma-Aldrich), and 5% PETA (pentaerythritol triacrylate, Sigma-Aldrich), with 1% (v/v) darocur 1173 photoinitiator. Each reagent has a distinct function as detailed. PEGDA primarily imparts hydrophilicity to the hydrophilic part of the Janus particles due to its linear ethylene glycol polymeric backbone. CEA further increases hydrophilicity and introduces a slight negative charge due to its carboxylic group, effectively minimizing any non-specific interactions. PETA ensures additional cross-linking density across the hydrophilic part due to three reactive sites, minimizing swelling and maintaining the intended cylindrical geometry. All monomers, along with the initiator, underwent N₂ purging for 1 hour before use. The hydrophobic solution was loaded into the PFPE micromold, the excess solution was removed by pipette suction, and then UV-irradiated in an N₂ atmosphere for 40 seconds to polymerize, creating the hydrophobic part of the Janus particles. Subsequently, the hydrophilic solution was loaded into the mold with the pre-formed hydrophobic part, excess solution removed, and further UV-irradiated in an N₂ atmosphere for 2.5 minutes, completing the formation of the Janus particles. Finally, the collection of the Janus particles from the PFPE mold was achieved through the application of acetone, which serves to gently release the particles without compromising their structural integrity, followed by washing with IPA several times. This methodical approach ensured the precise fabrication of Janus particles with distinct hydrophobic and hydrophilic parts.

DDDMA hydrophobic and PEG-CEA-PETA hydrophilic films were generated utilizing prepolymer solutions identical to those used for Janus particle preparation. Briefly, each solution was spin-coated onto 3-(Trimethoxysilyl) propyl methacrylate (TPM) pre-coated glass substrates and polymerized under N₂ through UV irradiation. Contact angles were measured using water (W), diiodomethane (DIM), and ethylene glycol (EG) as probe liquids, employing an optical tensiometer (Theta Lite, KSV Instruments, Helsinki, Finland) on each polymeric film.

Before loading the Janus particles into the capillary tube (0.2 mm × 2.0 mm ID, 3520S-050, Vitrocom), PEG silane (2-[Methoxy(Polyethyleneoxy)6-9Propyl] Trimethoxysilane (tech-90, SIM6492.7), Gelest) was coated to prevent nonspecific adhesion of the Janus particles to the tube wall. The capillary was filled with PEG silane using capillary action, followed by thermal incubation at 80°C in a convection oven for 5 minutes. Subsequently, the capillary was thoroughly washed with ethanol and deionized water multiple times, and any residue was removed by N₂ blowing. Next, anisotropic Janus particles with controlled hydrophobic and hydrophilic ratios (3:7, 5:5, 7:3) were initially dispersed in distilled water to a particle volume fraction of 0.004 (v/v, 0.4%), and each dispersion was carefully loaded individually into the PEG-coated capillary tubes to ensure homogeneity. Subsequently, a magnetic stir bar (D × L = 0.1 mm × 1 mm, kindly provided by NASA Glenn) was loaded into the capillary containing the dispersion. To minimize the risk of evaporation, both ends of the capillary tube, containing the particle solution, were securely sealed with the capillary wax (HR4-328, Hampton Research).

The capillaries filled with the Janus particles, prepared by the method described in 2.3., were placed on the horizontally balanced optical table, then the characteristics of the Janus particles and their behavior in gravity were observed on an inverted fluorescence microscope (TE2000, Nikon, Japan) equipped with a CCD camera (Coolsnap cf, Photometrics, USA). Image Pro software (Media Cybernetics, Maryland, USA) was used to obtain bright-field microscopy images of the Janus particle geometry, their time-dependent trajectory, and their self-assembled clusters as time increased.

This study is a part of the series of the colloidal experiment in the ISS, specifically the Advanced Colloids Experiment (ACE) Temperature Control-1 (ACE-T-1). For ACE-T-1 mission on ISS, the Janus particles with three different ratios (3:7, 5:5, 7:3) were loaded into each capillary according to the method described in 2.3., respectively. Then the capillaries were installed in a custom-engineered experimental module designed to establish temperature regulation and magnetic stirring, ensuring controlled experimental conditions. This module was subsequently transported to the ISS using a SpaceX spacecraft and installed in the Light Microscopy Module (LMM) on the ISS, specifically for conducting this space-based experiment. Remote operation of the space microscope from NASA’s GRC enabled real-time monitoring of the anisotropic particles’ behavior under microgravity. Before the initiation of imaging, a magnetic stir bar within the module was activated at 200 Hz for 1 hour, promoting the thorough mixing and uniform dispersion of the particles within the capillary tube. Imaging commences several minutes after the cessation of the stir bar’s rotation, allowing any flow induced by the magnetic stirring to subside, thus minimizing its influence on the particles’ dynamics. Briefly, we used a 2.5x objective to survey the sample and confirm its homogenization. Once homogenization is achieved, images were captured using 10x and 40x magnifications at 9 regions of interest (ROIs), which include the positions of the 5 thermistors and the areas between them. After capturing initial sets of images, a temperature gradient (the maximum gradient is Δ25°C) or a consistent temperature (capable range between 20°C to 60°C) was determined and applied for one of the predetermined samples to examine the temperature effect on the behavior of particles and their self-assembly. The temperature gradient value or a consistent temperature is controlled as needed throughout the mission. The images captured were transmitted back to NASA’s GRC for detailed analysis to elucidate the interactions and assembly behavior of the anisotropic Janus particles under microgravity conditions.

Image J software (http://rsb.info.nih.gov/ij/) was used to characterize the size of the hydrophobic and hydrophilic parts. We measured the dimension of the Janus particles using bright-field microscopic images that were focused on the equatorial (center) plane. In addition, the time-dependent displacement for the estimation of the two-dimensional (2D) translational diffusion coefficient (D_t) of the Janus particles was obtained from the trajectory of the particles analyzed via Image J integrated with PTA and Manual Tracking plugins. The estimated translational diffusion coefficient (D_t) was obtained individually from each data of gravity and microgravity experiment for comparison study.

To validate this, the hydrophobic interaction and hydrophilic repulsion between particles are approximately estimated by the theoretical thermodynamic free energy of adhesion (ΔG^ad). Briefly, the surface free energy of each hydrophobic and hydrophilic part is first calculated by using contact angles’ values of three different liquids, (Water (W), diiodomethane (DIM), and ethylene glycol (EG)), measured on each hydrophobic and hydrophilic film, and the van Oss-Chaudhury-Good (vOCG) equation (1) and equations (2, 3) (Vanoss et al., 1988). (1) λL(1+cosθ)=2γsLWγLLW+2γs+γL−+2γs−γL+ {\lambda _L}(1 + cos\theta ) = 2\sqrt {\gamma _s^{LW}\gamma _L^{LW}} + 2\sqrt {\gamma _s^ + \gamma _L^ -} + 2\sqrt {\gamma _s^ - \gamma _L^ +} (2) γsAB=2γs+γL− \gamma _s^{AB} = 2\sqrt {\gamma _s^ + \gamma _L^ -} (3) γstotal=γsLW+γsAB \gamma _s^{total} = \gamma _s^{LW} + \gamma _s^{AB}

Then, the interfacial free energies per unit contact area solid–liquid (γ₁₃, γ₂₃) and solid–solid (γ₁₂) interactions are determined using the surface free energy values (γ^tot, γ^LW, γ⁺, γ⁻) of the hydrophobic part, hydrophilic part, polar solvent, and equation (4) (Vanoss et al., 1988). (4) γij=γi+γj−2γiLWγjLW+2γi+γj−+2γi−γj+ {\gamma _{ij}} = {\gamma _i} + {\gamma _j} - 2\sqrt {\gamma _i^{LW}\gamma _j^{LW}} + 2\sqrt {\gamma _i^ + \gamma _j^ -} + 2\sqrt {\gamma _i^ - \gamma _j^ +}

Finally, we examined the free energy of adhesion (ΔG^ad) per unit contact area between the Janus particles (e.g., hydrophobic (γ₁) and hydrophilic (γ₂) parts) immersed in the assembly media (γ₃) by using the evaluated each interfacial free energy and equation (5, 6). (5) ΔG131ad=−2γ13 \Delta G_{131}^{ad} = - 2{\gamma _{13}} (6) ΔG232ad=−2γ23 \Delta G_{232}^{ad} = - 2{\gamma _{23}}

The measured contact angle and estimated free energy of adhesion is shown in Table 1. The distinct contact angles of water (CA=108.33°±2.22°) on the DDDMA polymeric films, clearly demonstrate it is hydrophobic, whereas the contact angle of an air bubble on the PEG-CEA-PETA film immersed in water (CA=35.95°±0.46°) is much lower, meaning it is hydrophilic. Evaluating the surface free energy components reveals that the apolar component (γ_s^LW) of the DDDMA surface is 36.15 mJ/m², while the Lewis acid and base components are negligibly small. This emphasizes the hydrophobic nature of the DDDMA part in the Janus particles, in line with the water contact angle results. Conversely, the PEG-CEA-PETA surface exhibits a surface free energy of 60.73 mJ/m² for the Lewis base component (γ_s⁻), with the Lewis acid component (γ_s⁺) being negligibly small, indicating the hydrophilicity and monopolarity of the PEG-CEA-PETA part in the Janus particles. The calculated free energy of adhesion (ΔG^ad) between hydrophobic DDDMA surfaces yields a highly negative value (−84.37), indicating a potential for spontaneous attraction through hydrophobic interaction. In contrast, the free energy of adhesion between hydrophilic PEG-CEA-PETA surfaces is significantly positive at +50.46, clearly indicating non-attraction, as detailed in Table 1. This non-attraction is likely attributed to the monopolar properties of the PEG-CEA-PETA surface, showcasing the ability to repel each other in polar solvents such as alcohol and water (Napper, 1983; Oss, 2006).

These results strongly support the anticipated precisely controlled interparticle interaction and their directed assembly of Janus particles through the combination of hydrophobic interaction and hydrophilic repulsion.

Next, the second design rule involved granting a degree of freedom for the continuous growth of self-assembled structures by finely tuning the geometry of the Janus particles. Beyond simple structures or small clusters such as dimers (N=2), constituted by interaction of a few Janus particles, we hypothesized that this continuous growth demands securing sufficient hydrophobic parts remaining even after the initiation of self-assembly. Furthermore, hydrophilic body should possess a minimal area to easily allow additional joining of the Janus particles to the initially formed assembled structures. For this, we designed the hydrophobic part of the Janus particles using convex-top geometry with a cylindrical body, which likely forms the self-assembled structure through minimal area interaction of the hydrophobic area and enabling its continuous growth without steric hindrance. To further enhance the degree of freedom for continuous growth, we hypothesized that convex-top Janus particles in the aqueous solution, varying in their ratios of hydrophilic to hydrophobic parts, are used for the experiment of self-assembly, and can form three distinct types of pairwise assemblies by hydrophobic interaction: tip-to-tip, tip to side, and side-to-side. Here, ‘tip’ refers to the small contact area of hydrophobic part, and ‘side’ refers to the lateral aspect of the hydrophobic part of the Janus particle, respectively.

Taken together, the two design rules for both interparticle force controlled via hydrophobic interaction and hydrophilic repulsion, and the degree of freedom for continuous growth affected by controlling hydrophobic and hydrophilic ratio with convex-top on the cylindrical body of the particles, we expected the behavior of the Janus particles and their self-assembly (FIG. 1). First, the thermodynamic principles guide particle interaction, with red representing hydrophobic parts and blue representing hydrophilic parts. A thermodynamically favorable configuration is achieved when the hydrophobic parts interact ΔG131ad<0 \Delta G_{131}^{ad} < 0 whereas a repulsive force predominates between hydrophilic parts ΔG232ad>0 \Delta G_{232}^{ad} > 0 (FIG. 1A). Once the initial assembled structure is formed, the ratio of hydrophobic and hydrophilic parts has significant potential to consecutively influence the degree of freedom for continuous growth. Presumably, the ratio of 3:7 may afford additional interaction with 1 (solid) or 2 (dash) free Janus particles, while the ratio of 5:5 may provide the freedom for interaction with 2 (solid) or 3 (dash) more Janus particles, and the ratio of 7:3 could allow interaction with 3 (solid) or 4 (dash) more Janus particles. Solid and dashed arrows represent higher or lower probabilities of interaction, respectively. This determination is based on the ratio between the hydrophobic and hydrophilic parts, the kinetic behavior of particles, and their thermodynamic equilibrium (FIG. 1B).

Figure 1.

Two crucial design principles for Janus particles based on chemical and geometrical anisotropy, and expectation on possible degree of freedom for continuous growth controlled by the ratio between hydrophobic and hydrophilic parts of the Janus particles. (A) Chemical anisotropy of the Janus particle: Precise control of interparticle interaction involving hydrophobic interaction and hydrophilic repulsion. (B) Geometrical anisotropy of the Janus particle: The hydrophobic (red) and hydrophilic (blue) ratio of the Janus particles possibly affects the interaction between initially assembled structures and free Janus particles, allowing different degrees of freedom for continuous growth during the self-assembly. For example, a 3:7 ratio allows interaction with 1 or 2 more Janus particles, 5:5 with 2 or 3 more, and 7:3 with 3 or 4 more. The additional geometrical anisotropy in the form of a convex-top design plays a role by enabling hydrophobic interactions with minimal contact area, thereby supporting continuous growth.

In alignment with the two critical design rules, the Janus particles, featuring a convex-top and a cylindrical body with precisely controlled hydrophilic and hydrophobic ratios, are produced through sequential micromolding. (Choi et al., 2010; Yeom et al., 2016). Briefly, a hydrophobic photocurable solution based on DDDMA is loaded into a PFPE micromold, then evaporation of diluent and subsequent irradiation of UV resulted in polymerization of formation of the hydrophobic part with a different ratio for the amount of the mixed diluent. Subsequently, a hydrophilic photocurable solution composed of PEG-DA, CEA, and PETA is loaded and polymerized via UV exposure, resulting in a convex-top Janus particle. The bright-field image shows reliable fabrication of the convex-top Janus particles with three different ratios via micromolding. The convex-top Janus particles are highly uniform in both geometry and size which is aligning precisely with the predesigned pattern of the PFPE micromold (FIG. 2A). The relatively different contrast in a Janus particle shown in the enlarged inset image of the FIG 2A is showing the boundary of two different parts: the hydrophobic part (dark) and hydrophilic part (bright) as designed. The controlled ratio between the hydrophobic and hydrophilic parts is a key geometrical anisotropy in our Janus particles, which plays a pivotal role in the ratio-dependent self-assembly, as elaborated-on in FIG. 1. Hence, precise characterization of this controlled ratio in the Janus particle is a critical prerequisite before proceeding with the valuable space experiments.

Figure 2.

Convex-top Janus particles with three different ratios (3:7, 5:5, 7:3) of hydrophobic and hydrophilic parts showing the geometrical anisotropy. (A) Schematic images and brightfield images showing uniformly controlled geometric aspects: the three different ratios (3:7, 5:5, 7:3) and the convex-top. In the inset images of the Janus particles, the hydrophobic parts are notably darker, while the hydrophilic parts appear brighter, highlighting the clear boundary between these distinct parts. (B) Quantitative analysis illustrating the geometrical anisotropy, ratios of the hydrophobic and hydrophilic parts of the convex-top Janus particles as designed (3:7, 5:5 and 7:3). Scale bars in all images are 30 μm.

Quantitative analysis clearly demonstrates the high uniformity in ratios, aligning closely with the targeted proportions of 3:7, 5:5, and 7:3 (FIG. 2B). These results clearly indicate that micromolding enables reliable preparation of the convex-top Janus particle with the sizes as well as the controlled geometry as designed.

Preceding the space experiments, behavior of the convex-top Janus particles with varying ratios (3:7, 3:3, 7:3) and their self-assembly in aqueous solution are investigated under the Earth’s gravitational forces for a comprehensive understanding of self-assembled structures and Brownian motion (FIG. 3). The Janus particles, characterized by their distinct hydrophobic and hydrophilic parts as chemical anisotropy, demonstrate directional interaction due to their inherent interaction forces between hydrophobic parts and the repulsive force between hydrophilic parts. FIG. 3A shows the representative self-assembled clusters and their type of pair configurations formed via self-assembly for each of the Janus particles (3:7, 5:5, 7:3).

Figure 3.

2D assembly of convex-top Janus particles under gravity. (A) The representative self-assembled 2D clusters and their type of pair configurations formed via self-assembly of each Janus particles (3:7, 5:5, 7:3). (B) Sequential time-lapse images showing the formation of the limited small dimer (N=2) staying at a similar position with a slight orientation change over time. This result demonstrates gravitational sedimentation severely interferes with the movement of particles as well as their self-assembly. (C) 2D mean squared displacement (MSD) of each of the convex-top Janus particles (3:7, 5:5, 7:3) in gravity. A slope (yellow) of the solid line divided by 4 yields the 2D translational diffusion coefficient of each convex-top Janus particle. Scale bars in the image showing the 7:3 Janus particle is 5 μm.

The optical images show the formation of the dimer (N=2) on a 2D planar plane through hydrophobic interactions between the convex-top Janus particles, suggesting that the reduction in surface-free energy is the predominant driving force behind the interactions in this particular geometric arrangement. Unexpectedly, the predominant configurations consist mainly of 2D dimers on a planar plane, regardless of the hydrophobic to hydrophilic ratio in the Janus particles. This observation might be a result of the limited time available for the Janus particles to undergo additional interparticle interactions, influenced by gravitational forces. In contrast, the ratio between the hydrophobic and hydrophilic parts of the Janus particles likely affects the pair configuration of the assembled clusters. The 3:7 Janus particles predominantly assemble in the tip-to-tip configuration (67%), with the remainder formed by the tip-to-side (33%) configuration. Interestingly, the 5:5 ratio exclusively shows the tip-to-side configuration (100%), while the 7:3 ratio exhibits the prevailing tip-to-side configuration (83%) with a minor fraction in the side-side configuration (17%). The different favorable configurations of the three different ratios (3:7, 5:5, 7:3) are presumably determined by a force balance between the hydrophobic interaction and hydrophilic repulsion of the Janus particle. As the particles approach to interact with each other, two forces act simultaneously: hydrophobic interactions minimize the unfavorable interface between the hydrophobic parts and water, whereas hydrophilic repulsion facilitates the formation of a favorable interface between the hydrophilic parts and water, along with repulsion between the particles. At this stage, the Janus particle configuration is governed by the ratio of their hydrophobic and hydrophilic parts. In the 3:7 ratio with a larger hydrophilic part, hydrophilic repulsion becomes dominant while hydrophobic interactions play a minimal role in particle orientation. The larger repulsive hydrophilic part favors aligning the hydrophilic parts of the particle spatiotemporally farthest away from each other to form a larger interface with water, allowing for the majority of the tip-to-tip configuration. In contrast, in the 7:3 ratio with a larger hydrophobic part, hydrophobic interactions become dominant while hydrophilic repulsion plays a minimal role in particle orientation. Considering the entire interface between Janus particles with water, a larger thermodynamically unfavorable interface is formed at the 7:3 ratio than at the 3:7 ratio, which needs to be minimized promptly. At this stage, the two configurations: tip-to-side and side-to-side, which minimize the larger interface with water, become dominant over tip-to-tip. However, the hydrophilic repulsion is not strong enough for reorientation, so the majority remains in the tip-to-side configuration, which is assumed to be the initial contact during particle interaction. In the case of a 5:5 ratio, the size of each part is identical, allowing a balanced contribution of hydrophobic interaction and hydrophilic repulsion to particle orientation. Hydrophobic interactions aim to reduce free energy by minimizing hydrophobic interfaces with water, while hydrophilic repulsion favors the formation of hydrophilic interfaces with water, but is not strong enough to keep the particles farthest away from each other. This balanced contribution of the two forces results in the exclusive formation of the tip-to-side configuration.

These configuration differences even in gravity provide promising expectations that the degree of freedom can potentially be altered depending on the ratio of the hydrophobic and hydrophilic parts of the Janus particles. The analysis of time-lapse images (t₁, t₂, t₃) shows the progressive stages of particle interaction (FIG. 3B). However, a significant number of the Janus particles sediment before larger cluster formation, primarily as a consequence of gravitational forces.

We further estimate the gravitational Péclet number (Pe), which is a dimensionless quantity describing the ratio of the time of the Brownian diffusion (τ_B) against that of sedimentation (τs) of particles, using the equation (7) (Lee and Furst, 2006; Whitmer and Luijten, 2011). (7) Pe=τBτs=σ2lg=πσ4Δρg12kBT Pe = {{{\tau _B}} \over {{\tau _s}}} = {\sigma \over {2{l_g}}} = {{\pi {\sigma ^4}\Delta \rho g} \over {12{k_B}T}}

Where σ is the diameter of the particle, l_g is the gravitational length, Δρ = |ρ_p – ρ_s| is the density difference between the particle and the solvent, g is the gravitational acceleration, and k_BT is the thermal energy. The calculated Pe on Earth is 77.083 (Pe >> 1), indicating that gravity significantly influences the particles, with sedimentation prevailing over Brownian motion. In practical terms, this means that particles descend so rapidly due to gravity that there is inadequate time and velocities for them to fully explore the phase space of positions essential for thermodynamic assembly processes.

In addition, it’s important to note that the gravitational Pe reflects the characteristics of individual particles, meaning the impact of gravity can be substantially different at assembled clusters. For instance, once the Janus particles form self-assembled clusters, the relevant buoyant force is not that of the single Janus particle but rather the force experienced by assembled clusters consisting of multiple numbers of the Janus particles.

Subsequently, evaluating the translational diffusion coefficient of convex-top Janus particles in a gravitational field and conducting a comparative study against the theoretically estimated diffusion coefficient is meaningful for a more in-depth understanding of particle behavior. To achieve it, 2D locational displacement (x, y) of a single particle is obtained by tracking the single particle, and all displacements in both the x- (e.g. Δx₁=x₁− x₀) and y-direction (e.g. Δy₁=y₁− y₁) derived from the 2D trajectory as a function of time are obtained. Next, squared total displacement (Δr²) is calculated by the squaring and summation of each displacement (Δr²=Δx²+Δy²). Then, MSD (〈(Δr²)〉) is computed by averaging Δr², which is obtained from 10 particles’ tracking analysis.

In FIG. 3C, the MSD over time of each convex-top Janus particle (3:7, 5:5, 7:3) is shown. In 2D, the time-dependent MSD can be expressed as 〈(Δr²)〉=4D_t·Δt. By dividing the slope obtained from linear regression by 4, the translational diffusion coefficients (D_t) for each convex-top Janus particle are derived. The estimated diffusion coefficient describes the behavior of particles diffusing through a medium, which has a high correlation with the movement and interaction of particles and collision frequency, thereby understanding self-assembly. The calculated translational diffusion coefficients (D_t) for the convex-top Janus particles are 4.523×10⁻² μm²/s (3:7 ratio), 3.450×10⁻² μm²/s (5:5 ratio), and 3.773×10⁻² μm²/s (7:3 ratio), respectively. Unfortunately, there is no significant correlation between the ratio of the Janus particles and the diffusion coefficient, which may be hindered due to gravity. For comparison, the theoretical diffusion coefficient is computed using the bead shell-based cylinder model equation (8) (Carrasco and de la Torre, 1999; Delatorre and Bloomfield, 1981; Ortega and de la Torre, 2003; Tirado et al., 1984). Here, η represents the fluid viscosity, while L and d denote the length and diameter, respectively, and C_t stands for a numerical constant. (8) (Dt=kBT3πηL(ln(L/d)+Ct)) ({D_t} = {{{k_B}T} \over {3\pi \eta L}}(ln(L/d) + {C_t}))

The convex-top Janus particle exhibits exceptional chemical and geometrical anisotropy, so pinpointing an ideal theoretical model for Janus particles proves challenging. To the best of our knowledge, the bead shell-based cylinder model provides a valuable initial framework for comprehending particle behavior. The theoretically computed diffusion coefficient as 9.658×10⁻² μm²/s reveals that the Brownian diffusion of the Janus particle in gravity is approximately 2 or 3 times less than the theoretical model. Therefore, the constrained motion of Janus particles, caused by gravitational stress, is believed to restrict additional freedom in interparticle interaction, consequently hampering the formation of larger or more intricate clusters.

Next, we perform the ACE-T-1 space experiment to investigate promising design parameters for enabling selective interactions with directional specificity to build complex structures of significance. Utilizing the LMM at ISS, our research investigates the controllability of Janus particles to pioneer the construction of three-dimensional (3D) structures. The optical images in FIG. 4 reveal the distribution of 3D clusters formed by convex-top Janus particles through hydrophobic interactions and hydrophilic repulsions under microgravitational conditions. As designed, the chemical anisotropy comprises hydrophobicity and hydrophilicity to effectively enable directional interaction of the Janus particles. These 3D clusters originate from a combination of three configurations, tip-to-tip, tip-to-side, and side-to-side, within the hydrophobic parts of particles characterized by the convex-top geometry. Unlike the bidirectional interactions on the planar plane of Earth that result in 2D assemblies, the clustering of convex-top Janus particles exhibits multiaxial orientations in 3D space. This complexity renders it statistically challenging to classify these assemblies into a specific configuration. (FIG. 4A). Observation of these 3D clusters, which consist of the same number of particles but exhibit distinct spatial orientations, allows us to deduce that upon initial contact, two Janus particles can readily reorient to establish energetically more favorable assemblies. This adaptability in altering their interparticle configurations sharply contrasts with the behavior of chemically homogeneous cylinders, which tend to form robust capillary bridges between their planar end surfaces, resulting in the formation of linear chains. This distinct difference is attributed to the unique interactive dynamics of Janus particles compared to their chemically uniform counterparts, implying their potential for creating complex, energy-efficient structures.

Figure 4.

3D assembly of convex-top Janus particles in microgravity. (A) The representative self-assembled 3D dimers (N=2) and their various multiaxial orientation pair configurations formed via self-assembly of each of the Janus particles (3:7, 5:5, 7:3). (B) Various forms of self-assembled 3D clusters such as spherical-like, densely packed, and linearly grown clusters that are thermodynamically favorable and considered an equilibrium state. (C) Sequential time-lapse images showing an initially assembled dimer (N=2) of the convex-top Janus particle and its continuous growth over time to tetramer (N=4). (D) 2D MSD of each convex-top Janus particle (3:7, 5:5, 7:3) in microgravity. A slope (yellow) of the solid line divided by 4 yields a 2D translational diffusion coefficient (D_t) of each convex-top Janus particle.

FIG. 4B demonstrates representative forms of assembled 3D clusters under microgravity using each of the Janus particles. Unlike the results on Earth, it is observed that the cluster size and form are significantly influenced by the hydrophobic part. Typically, the Janus particles with small hydrophobic parts self-assemble into seed clusters containing mainly dimers (N=2) and trimers (N=3) or a few of multimers (N≥4) in spherical-like form. In contrast, particles with large hydrophobic parts self-assemble into densely packed or linearly grown forms composed mainly of more multiparticles (N≥4). T h i s result can likely be attributed to the combined effects of geometry and chemical properties.

Through localized contact attributed to the convex-top geometry, primarily in a tip-to-tip configuration, small seed clusters are initially formed. In this state, there are still many remaining hydrophobic areas in the pre-formed small clusters that can potentially interact with other individual Janus particles, which is a favorable environment for continuous cluster growth. Then, the equilibrium states exhibited by the final assembled clusters are mainly determined by the ratios of hydrophobic and hydrophilic parts. In particular, the 7:3 Janus particle, characterized by the largest hydrophobic part, demonstrates the highest degree of freedom for continuous growth. This is attributed to the substantial residual hydrophobic part on the initial seed cluster, enabling interactions with subsequently joining particles from any axial orientations. As a result, this promotes the formation of clusters with linear growth or dense packing. In the case of the 3:7 ratio, which is characterized by the smallest hydrophobic part, the predominant form of the self-assembled structure consists of loosely arranged, 3D spherical-like clusters. Although the initial seed cluster formed by 3:7 Janus particles possess a relatively smaller degree of freedom for continuous growth due to the smallest hydrophobic part, there is still remaining space on the hydrophobic part, making it thermodynamically favorable for continuous growth. However, as the growth proceeds, the remaining hydrophobic space diminishes, necessitating specifically oriented access for the next joining particles. Moreover, relatively stronger hydrophilic repulsion from the largest hydrophilic part coexists which supports the center orientation of particles more precisely. This leads to the enforcement of center-oriented hydrophobic parts in self-assembled clusters, resulting in loosely arranged 3D spherical-like clusters. The 5:5 Janus particle, with an intermediate proportion of hydrophobic parts between 3:7 and 7:3, exhibits characteristics of both, thereby forming all types of structures including linearly grown or loosely arranged spherical-like or densely packed clusters. In other words, the results described in FIG. 4B indicate that the final structure of the assembled clusters can be designed by manipulating the geometrical aspect, specifically the ratios of hydrophobic and hydrophilic parts of the Janus particles.

The optical images obtained under microgravity align with the 3D illustrations in FIG. 4C, depicting the temporal evolution and spatial arrangement of 3:7 convex-top Janus particle assembly. In the unique microgravity environment devoid of gravitational sedimentation (estimated Pe=7.78×10⁻⁵), the Janus particles are actively and freely exploring the full phase space of positions, allowing the formation of the 3D cluster. Once aligned with thermodynamically favorable orientations, they likely maintain their initial configuration stability until interacting with other particles. These unique assembly systems allow for the stable formation of seed clusters (N=2) with spatially diverse orientations that maintain a high degree of freedom for continued growth, resulting in a variety of 3D clusters ranging from those that are spherical-like, densely packed, to linearly grown forms.

Moreover, the unique environment provided by microgravity uncovers the hidden fundamental kinetic information of the Janus particles. It is important to note that the fascinating microgravity conditions allow the Janus particles to actively explore 3D spaces, which is highly promising. However, we encounter challenges in analyzing their 3D trajectories. The particle selected for tracking frequently and completely moves out of the microscope’s focal plane, which leads to analytical challenges in obtaining Z-axis displacement once the particle disappears. Consequently, conducting our tracking and analysis solely in 2D trajectories is inevitable.

FIG. 4D shows translational diffusion coefficients (D_t) of each of the Janus particles (3:7, 5:5, 7:3) under microgravity, as evaluated by analyzing 2D trajectory. Compared to the diffusion coefficient obtained in gravity, we confirm that each diffusion coefficient of the particles for 3:7 (8.618×10⁻² μm²/s), 5:5 (9.785×10⁻² μm²/s), and 7:3 (9.378×10⁻² μm²/s) closely match the theoretically estimated diffusion coefficient (D_t=9.658×10⁻² μm²/s) for a bead shell-based cylinder. Unfortunately, the effect and correlation of the ratio of hydrophobic and hydrophilic parts on the diffusion coefficient remained unclear. This may be due to several differences between the Janus particles and the theoretical model. For example, the theoretical model is based on a cylinder with flat-ends at both sides, but the Janus particle is a cylinder with a convex-top and a flat-end. In addition, the theoretical value is considered for the aspect ratio (AR)>2, which is quite larger than the AR of the sample (AR≈1.67). Furthermore, the Janus particles are chemically anisotropic and composed of both hydrophilicity and hydrophobicity.

Overall, this experimental result strongly suggests that the absence of gravity permits a more accurate characterization of the particles’ behavior, unaffected by the sedimentation that occurs on Earth. This could potentially lead to the discovery of new self-assembling structures and mechanisms, contributing to the advancement of material science and engineering.

Next, ACE-T-1 mission’s notable advancement over previous NASA colloidal science missions is the ability to control the temperature of sample cells. This capability allows for an in-depth exploration of Janus particles’ behavior in microgravity under precisely controlled temperatures, encompassing particle kinetics, thermal interactions, and self-assembly. For this purpose, we conduct a detailed analysis of cluster formation among Janus particles with a hydrophobic-to-hydrophilic ratio of 7:3 and quantitatively evaluate the motion of the Janus particles at different temperatures. ACE-T-1 utilizes capillary cells, and one set of three capillary cells is provided for each sampling module. On a stage with temperature control, three capillary cells can be positioned so that a uniform temperature can be set for the entire sample, or a temperature gradient can be placed along the length of the capillaries (typical range: 20–60°C, 10°C temperature differential over the length of the capillary) or between the bottom and top of a capillary in case an oil immersion objective is used as a cold sink (which remains close to ambient temperature, around 20°C).

Figure 5.

Temperature-dependent particle kinetics and their self-assembly. (A) The MSD of the Janus particles at each temperature (T≈20°C and T=60°C). A slope (red or blue) of the solid line divided by 4 yields the 2D translational diffusion coefficient of the 7:3 convex-top Janus particle at each temperature. The inset images show distinct 3D structures formed via self-assembly at each temperature: complex 3D clusters with smaller contact areas at 60°C and less complex 3D clusters with larger contact areas at 20°C. (B) Suggested behavior, interaction of the Janus particles, and 3D complex cluster formation at each temperature.

FIG. 5 presents a pivotal discovery from the ISS experiments regarding the temperature-dependent dynamics of anisotropic Janus particle assemblies. FIG. 5A shows the MSD analyzed over increasing lag times (Δt) for 7:3 Janus particles, providing quantitative insights into particle mobility at 20°C and 60°C. At 60°C, the particles exhibit significantly higher MSD, indicating enhanced kinetics of the particles, compared to their behavior at 20°C. Interestingly, the inset images in FIG. 5A reveal that the 3D structures formed by self-assembly at each temperature differ markedly: complex 3D clusters with smaller contact areas form at 60°C, while less complex 3D clusters with larger contact areas form at 20°C.

FIG. 5B suggests that the variation in 3D assemblies at different temperatures is presumably due to variations in particle dynamics and interparticle forces. The formation of complex 3D clusters with smaller contact areas at 60°C is likely attributed to increased particle kinetics and hydrophobic interactions. Elevated temperatures enhance particle kinetics, increasing interaction opportunities between particles and pre-assembled seed clusters. In the case of hydrophobic interaction, it is known to increase as temperature increases (Baldwin, 1986; Claesson et al., 1986; Kegel and van der Schoot, 2004) because of its endothermic nature (Némethy and Scheraga, 1962). Therefore, we hypothesize that elevated temperature imparts strong enough hydrophobic interaction to form self-assembly with a smaller contact area between hydrophobic parts. This cluster still provides many leftover areas for continuous growth, allowing additional interaction with kinetically increased particles. However, at a relatively lower temperature at 20°C, relatively less kinetics of particles and weaker hydrophobic interactions necessitate larger contact areas for equilibrium, resulting in less complex 3D assemblies.

The quantitative analysis in Table 2 provides the 2D translational diffusion coefficient, indicating particles’ mobility and the theoretically estimated kinetics of interparticle interaction at each temperature. First, the translational diffusion coefficient (D_t) of the 7:3 hydrophobic-to-hydrophilic ratio of the Janus particles at 60°C is 2.28 times greater than at 20°C, evidently showing that the Janus particles’ random motion is quite active at higher temperatures. Next, the collision frequency between the Janus particles at two different temperatures (≈20°C and 60°C) is approximately assessed by employing the Smoluchowski equation (9) (Gregory, 2005; Wang et al., 2012) for estimation of the increased kinetics of the Janus particles. (9) J≈4π(ri+rj)(Di+Dj)NiNj J \approx 4\pi ({r_i} + {r_j})({D_i} + {D_j}){N_i}{N_j}

Here, r is radius of the particles i and j (here we assumed the Janus particles as spheres for approximation), D is the translational diffusion coefficient of the particles i and j, and N is number of particles i and j per unit volume, respectively. Basically, the 7:3 convex-top Janus particles in microgravity collide 2.5 times more frequently (0.133 sec⁻¹) than under gravity (0.054 sec⁻¹). At a higher temperature (=60°C), the collision frequency further increases to 2.3 times (0.304 sec⁻¹). Notably, collisions occur every 3.289 seconds at 60°C, while the collision time extends to 7.518 seconds at lower temperature (≈20°C). The elevated collision frequency at a higher temperature enhances interaction kinetics, leading to the formation of larger and more complex 3D clusters, whereas smaller clusters emerge at a lower temperature. This temperature-mediated modulation of particle dynamics and assembly complexity is of considerable interest. It suggests a methodology for controlling the self-assembly process of colloidal particles by adjusting the ambient temperature, thereby designing the final structures’ complexity. The findings prove that by harnessing the thermodynamic parameters, it may be possible to engineer novel materials with customized properties, utilizing the unique conditions of microgravity and controlled temperature to steer the self-assembly processes of colloidal systems.

NASA’s GRC, with its extensive experience and professional expertise in remote operations, has been integral to our mission operation. Their proficiency in remote control and data transfer processes has enabled the seamless operation of experimental parameter controls and data acquisition. The significant findings from our mission, such as Janus particles’ movement, orientation, kinetics, and self-assembly in space without the masking effect of gravity, have enhanced our understanding of the fundamental behavior of anisotropic particles. It is promising to enrich colloidal and advanced materials science research with novel applications and functionalities in an extraterrestrial environment.