Hostname: page-component-7b7567c8c-fvjsw Total loading time: 0 Render date: 2026-11-16T10:33:42.321Z Has data issue: false hasContentIssue false

Artificial chemotaxis of phoretic swimmers: instantaneous and long-time behaviour

Published online by Cambridge University Press:  12 October 2018

Maria Tătulea-Codrean
Affiliation:
Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Wilberforce Road, Cambridge CB3 0WA, UK
Eric Lauga*
Affiliation:
Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Wilberforce Road, Cambridge CB3 0WA, UK
*
†Email address for correspondence: e.lauga@damtp.cam.ac.uk

Abstract

Phoretic swimmers are a class of artificial active particles that has received significant attention in recent years. By making use of self-generated gradients (e.g. in temperature, electric potential or some chemical product) phoretic swimmers are capable of self-propulsion without the complications of mobile body parts or a controlled external field. Focusing on diffusiophoresis, we quantify in this paper the mechanisms through which phoretic particles may achieve chemotaxis, both at the individual and the non-interacting population level. We first derive a fully analytical law for the instantaneous propulsion and orientation of a phoretic swimmer with general axisymmetric surface properties, in the limit of zero Péclet number and small Damköhler number. We then apply our results to the case of a Janus sphere, one of the most common designs of phoretic swimmers used in experimental studies. We next put forward a novel application of generalised Taylor dispersion theory in order to characterise the long-time behaviour of a population of non-interacting phoretic swimmers. We compare our theoretical results with numerical simulations for the mean drift and anisotropic diffusion of phoretic swimmers in chemical gradients. Our results will help inform the design of phoretic swimmers in future experimental applications.

Information

Type
JFM Papers
Copyright
© 2018 Cambridge University Press 
Figure 0

Figure 1.Sketch of the problem. (a) Axisymmetric active particle with axis of symmetry along $z$. The surface properties (surface activity $\unicode[STIX]{x1D70E}$, substrate mobility $\unicode[STIX]{x1D707}_{s}$ and product mobility $\unicode[STIX]{x1D707}_{p}$) are functions of the polar angle $\unicode[STIX]{x1D703}$ alone. (b) Active particle placed in a uniform substrate gradient, $\unicode[STIX]{x1D735}s^{\infty }$, which lies in the $yz$-plane. The presence of the background gradient, together with the catalytic coating of the swimmer, induce local tangential chemical gradients on the surface of the sphere (light grey arrows) which give rise to an apparent fluid slip velocity (white arrows) leading to translation (linear velocity $\boldsymbol{V}$) and rotation (angular velocity $\unicode[STIX]{x1D74E}$) of the particle.

Figure 1

Figure 2.Schematic representation of a Janus spherical swimmer with surface properties $\{\unicode[STIX]{x1D70E}^{U},\unicode[STIX]{x1D707}_{s}^{U},\unicode[STIX]{x1D707}_{p}^{U}\}$ on the upper hemisphere and $\{\unicode[STIX]{x1D70E}^{L},\unicode[STIX]{x1D707}_{s}^{L},\unicode[STIX]{x1D707}_{p}^{L}\}$ on the lower hemisphere. The linear and angular velocities of this phoretic swimmer are given by (2.50)–(2.53).

Figure 2

Figure 3.Comparison of the distribution of swimmer orientations obtained from numerical simulations with the predictions of the continuum model. (a) Histograms of final orientations for 10 000 swimmers whose initial orientations were chosen from a uniform distribution, after being evolved in time for 500 time steps of size $\unicode[STIX]{x0394}t=10^{-2}D_{r}^{-1}$. The red solid line corresponds to the analytical expression from (3.21) for the steady-state distribution of swimmer orientations. (b) Sample of 1000 swimmer orientations chosen from the steady-state distribution. Each swimmer orientation corresponds to a unit vector which is represented here as a blue dot on the unit sphere. From top to bottom the rotational Péclet number is $\unicode[STIX]{x1D706}=0.1,3,10$.

Figure 3

Figure 4.Time evolution of 200 independent trajectories, all starting from the origin at $t=0$. Each snapshot is taken after a time interval of $3D_{r}^{-1}$, and the chemical gradient is along the horizontal axis. The large dots are placed at regular distances of $3V_{s}D_{r}^{-1}$ from the origin which confirms that the cloud of phoretic swimmers drifts along the direction of the chemical gradient at the mean swimming velocity ($V_{s}$), as predicted by the continuum model. This simulation corresponds to parameters $\unicode[STIX]{x1D706}=5$ and $\unicode[STIX]{x1D707}=3$, and the time step used was $\unicode[STIX]{x0394}t=10^{-2}D_{r}^{-1}$.

Figure 4

Figure 5.Comparison of the values obtained from numerical simulations against those predicted by GTD theory for the dependence of component $D_{\bot }$ of the active diffusivity tensor on $\unicode[STIX]{x1D706}$. The simulations were run with 100 samples of 1000 swimmers using a time step $\unicode[STIX]{x0394}t=10^{-2}D_{r}^{-1}$ over a period $2D_{r}^{-1}$. Each sample was used to calculate one value of the diffusivity $D_{\bot }$, and then averaged to obtain one data point. The error bars represent one standard deviation amongst the 100 values obtained for the diffusivity. The solid lines represent the predictions of the continuum model from (3.32)–(3.35) and (3.38)–(3.39). In all plots the diffusivity is non-dimensionalised by $U^{2}D_{r}^{-1}$. The four graphs correspond to parameter values $\unicode[STIX]{x1D707}=0$, $\unicode[STIX]{x1D707}=1$, $\unicode[STIX]{x1D707}=5$ and $\unicode[STIX]{x1D707}=10$.

Figure 5

Figure 6.Iso-values of the diffusivities $D_{\bot }$ and $D_{\Vert }$ computed using GTD theory in the $(\unicode[STIX]{x1D706},\unicode[STIX]{x1D707})$ plane. Both diffusivities are normalised using the mean swimming velocity and the time scale of rotational diffusion, i.e. $D\mapsto D/(V_{s}^{2}D_{r}^{-1})$. These contour plots are obtained with a value of $\unicode[STIX]{x1D708}=1.5$ for the direct chemotactic index, but there are no qualitative changes as $\unicode[STIX]{x1D708}$ is varied (not shown). We observe consistent decay of both diffusivities as $\unicode[STIX]{x1D706}\rightarrow \infty$ at a fixed value of $\unicode[STIX]{x1D707}$, as well as a saddle point in diffusivity for $\unicode[STIX]{x1D706},\unicode[STIX]{x1D707}=O(1)$.

Figure 6

Figure 7.The ratio $D_{\bot }/D_{\Vert }$ computed using GTD theory in the $(\unicode[STIX]{x1D706},\unicode[STIX]{x1D707})$ plane. As expected, the contour $D_{\bot }/D_{\Vert }=1$ passes through the origin, because the system reverts to isotropic diffusion in the limit $\unicode[STIX]{x1D706},\unicode[STIX]{x1D707}\rightarrow 0$. Increasing $\unicode[STIX]{x1D707}$ above this contour leads to a ratio $D_{\bot }/D_{\Vert }<1$, meaning that indirect chemotactic sedimentation enhances diffusion along the chemical gradient relative to diffusion in the normal plane. On the other hand, increasing $\unicode[STIX]{x1D706}$ to the right of this contour leads to a ratio $D_{\bot }/D_{\Vert }>1$, meaning that chemotactic alignment suppresses diffusion along the chemical gradient relative to diffusion in the normal plane.