Programmable Hydrodynamics of Active Particles

Abstract

Self-propelled microparticles create flow fields that determine how they interact with surfaces, external flows, and each other. These flow fields fall into distinct classes--pushers, pullers, and neutral swimmers--each exhibiting fundamentally different collective behaviors. In all existing synthetic systems, this hydrodynamic character is permanently set during fabrication, making it impossible to explore how adaptive switching between these classes might enable new functions or emergent phenomena. Here we demonstrate that the hydrodynamic character of a microswimmer can be programmed and switched on demand. Using patterned laser heating of surface-bound nanoparticles, we create tailored temperature gradients that drive controllable boundary flows at the particle surface. By changing the illumination pattern in real time, we dynamically transform the swimmers flow field continuously tuning from pusher to puller, while the particle continues to swim. Flow measurements confirm quantitative agreement with theory and allow us to simultaneously track how symmetry, power consumption, and efficiency change across modes. This control over hydrodynamic modes opens experimental access to questions that have remained largely theoretical: How do adaptive swimmers respond to crowding or confinement? Can mixtures with tunable pusher-puller ratios reveal new collective states? Our approach provides a platform to address these questions and explore the morphological developments of active matter systems under external physical constraints.

Figures (4)

• Figure 1: Experimental realization of hydrodynamic reconfigurationa Reconfiguring hydrodynamic flows of a microswimmer is achieved by selectively illuminating different regions on the particle's surface via a spatial light modulator (SLM). b The microswimmer is a melamine sphere with radius $a = 2.2\,$µm that is uniformly covered with gold nanoparticles that behave as heat sources when illuminated with laser light at $532\,$nm. The presented structured light pattern induces a heating at the equator of the particle at $\theta = \pi$ which will lead to self-propulsion $U$ along $z$. The angle $\theta$ is defined as the polar angle to the positive z-axis across the surface of the microswimmer. c All heated nanoparticles create a tangential temperature gradient which generates a thermo-osmotic boundary flow $u_\theta$ indicated by the black arrows. This boundary flow creates a hydrodynamic flow field which is characteristic for the selected swimming mode and resembles here a force dipole field.
• Figure 2: Surface temperature control by precise illumination patternsa,e We precisely shape the laser field using a spatial light modulator (SLM) to generate distinct illumination patterns, as shown here for a pusher with $\beta = -1$ and a neutral squirmer with $\beta = 0$. The measured intensity line profiles (green) agree well with the theoretical predictions from Eq. \ref{['eq:q_modes']} (black), demonstrating accurate reproduction of the expected profiles. We obtain the microscopy images by exploiting the fluorescence of a dye at $\lambda = 532\,$nm. b,f The phase boundary of a liquid crystal forms when heating the microswimmer above the critical temperature of $308\,$K, revealing the asymmetric temperature profile on the surface of a swimmer, presented here for $\beta =-1$ and $\beta = 0$. The isotherms are highlighted by dashed lines defining the radius $r_\mathrm{exp}(\theta)$. c,g The angular dependence of $r_\mathrm{exp}(\theta)$, measured for different laser powers $P$ in the sample plane, confirm the precise control of the surface illumination. d,h The asymmetry of the isotherms ($r_\mathrm{exp}(\theta)$) gives information about the heat source density $q(\theta)$ on the particle's surface which can be used to reconstruct the surface temperature difference $\Delta T_\mathrm{surf}(\theta)$ as shown here for different heating powers $P$. Error bars represent the standard error of the mean calculated from multiple measurements.
• Figure 3: Reconfigurable propulsion modes through hydrodynamic flow fieldsa Experimental flow fields for various swimming modes $\beta$, visualized using particle image velocimetry (PIV). The motion of tracer particles is represented as streamlines, with the red arrow indicating the propulsion direction. b Simulated flow fields are presented. The excellent agreement between experimental and simulated flow fields demonstrates the precise control over boundary flows achieved in the experiments, enabling reconfigurable propulsion modes through controlled variation of the illumination patterns. c,d,e While the normalized $x$-component of the velocity $u_x/u_{x,\mathrm{max}}$ is symmetric around $x$, the $z$-component $u_z/u_{z,\mathrm{max}}$ (f,g,h) indicates the polar asymmetry of the hydrodynamic flow fields. The measured velocity components (dotted lines) closely follow the decay of simulated velocity components (solid lines), demonstrating quantitative agreement between experiment and simulation. The swimmer's surface is marked with yellow color. Error bars represent the standard error of the mean calculated from multiple measurements.
• Figure 4: Swimming efficienciesa The trajectory of an active particle undergoing sequential transitions between distinct propulsion modes — from neutral squirmer (red) to pusher (green), and subsequently to puller (blue) — demonstrates the capability for instantaneous reconfiguration. b Mean squared displacement (MSD) as a function of lag time $\tau$ for a neutral squirmer ($\beta = 0$), puller ($\beta = 1$), and pusher ($\beta = -1$) at the same laser power of $P = 4.7\, mW$ scales with $\tau^2$, indicating ballistic motion of the active particle. c The velocity of the swimmers increases linearly with laser power $P$. The neutral squirmer reaches velocity values three times larger than the pusher. The dashed lines show fits of the form $U = mP$. d The absorbed power $P_\mathrm{abs}$ as a function of squirmer parameter $\beta$ for constant $q_1=- 0.4\,µW \, (µ m)^{-1}$ indicates that the pusher with $\beta \approx -1$ exhibits the lowest absorbed power. e The experimentally determined efficiencies (symbols) for different swimming modes align well with the theoretical model (dashed line), demonstrating that the pusher with $\beta\approx -1$ is the most efficient swimmer and further validating the control and characterization of the propulsion modes. The non-linear and linear regimes are indicated by the gray and white backgrounds, respectively. Error bars represent the standard error of the mean calculated from multiple measurements.