Selfdiffusiophoretic Janus colloids

Abstract

Moving through a fluid is a common experience for all humans. Even if one is not a skilled swimmer any kind of paddling motion will propel you forward. The motion is based on transfer of momentum from the swimmer to the fluid. However, there are regimes, described by low Reynolds numbers, where this kind of propulsion does not work. For small length scales and sufficiently low velocities a fluid behaves similar to honey. This means, every kind of motion is immediately damped if it is not sustained by a permeant external force. For physics, this regime is of particular interest as many biological systems like bacteria or colloids are faced by these conditions. It was first pointed out in the seminal work of Eric Purcell [1] that in order to propel at low Reynolds numbers a permanent non-time-reversal motion has to be kept up, as for example the power and recovery strokes of bacterial cilia [2]. It inspired a lot of experimental work to build up low Reynolds number swimmers based on this principle [3]. They could e.g. be used in microfluidic devices [4] or as drug carriers in the human body. However, in order to let an artificial swimmer evolve through a set of configurations in a controlled way, some kind of external mechanism has to be applied. In most applications, a magnetic field is used to derive this effect. This can be a drawback if these swimmers have to explore complex environments. In the recent years, alternative approaches based on the motion of swimmers in external gradients, especially in particle gradients, received considerable interest. The particle gradient approach has been inspired by the motion of bacteria in gradients of nutrients, called chemotaxis. This mechanism, however, relies on a rather complicated internal machinery to reorientate the bacterium. An artificial swimmer has to be based on a more simple physical principle in order to guarantee highly controllable motion. One of the most promising candidates is diffusiophoresis. It explains the motion of a colloid in an external particle gradient based on an active process at the interface between the fluid and the solid [5]. Recently, this approach has been extended to so called selfdiffusiophoretic swimmers [6, 7] which are not driven by an external gradient but produce the gradient themselves. Even though there are first analytical approaches trying to describe this phenomenom [8, 9], due to the coupling of low Reynolds number hydrodynamics to the evolution of solute particle gradients it is difficult to obtain a complete description. An alternative are simulations which are able to capture both contributions. Classically, two approaches are used to simulate low Reynolds number hydrodynamics. The first one tries to mimic the effective force due to the fluid on the embedded objects, like for example for Stokesian Dynamic simulations [10]. The second explicitly models the fluid and its interaction with relevant boundary conditions. The former is a veritable tool to address e.g. the effect of hydrodynamics on the relaxation of polymers but it is difficult to implement more complex fluids as in the case of selfdiffusiophoretic swimmers. Therefore, the second approach is used for the simulations in this work. Most of the fluid models mimic a fluid in which thermal fluctuations can be neglected. This is a sufficient assumption if the motion of the fluid is