Bio-inspired Magnetic Systems: Controlled Swimming, Fluid Pumps, and Collective Behaviour

Abstract

This thesis details the original experimental investigations of magnetically actuated and controlled microscopic systems enabling a range of actions at low Reynolds number. From millimetre-robots and self-propelled swimmers to microfluidic and lab-on-a-chip technology applications. The main theme throughout the thesis is that the systems reply on the interactions between magnetic and elastic components. Scientists often take inspiration from nature for many aspects of science. Millimetre to micrometre machines are no exception to this. Nature demonstrates how soft materials can be used to deform in a manner to create actuation at the microscale in biological environments. Nature also shows the effectiveness of using beating tails known as flagella and the apparent enhancements in flow speeds of collective motion. To begin with, a swimmer comprised of two ferromagnetic particles coupled together with an elastic link (the two-ferromagnetic particle swimmer), was fabricated. The system was created to mimic the swimming mechanism seen by eukaryotic cells, in which these cells rely on morphological changes which allows them to propel resulting in approximate speeds of up to 2 body lengths per second. The aim of this system was to create a net motion and control the direction of propagation by manipulating the external magnetic field parameters. It was shown that the direction of swimming has a dependence on both the frequency and amplitude of the applied external magnetic field. A key factor discovered was that the influence of a small bias field, in this case, the Earth’s magnetic field (100 orders of magnitude smaller than the external magnetic field) resulted in robust control over the speed (resulting in typical swimming speeds of 4 body lengths per second) and direction of propulsion. Following this work, swimmers with a hard ferromagnetic head attached to an elastic tail (the torque driven ferromagnetic swimmer) were investigated. These systems were created to be analogous to the beating flagella of many natural microscopic swimmers, two examples would be sperm cells and chlamydomonas cells. These biological cells have typical speeds of 10s of body lengths per second. The main focus of this investigation was to understand how the tail length affects the swimming performance. An important observation was that there is an obvious length tail (5.7 times the head length) at which the swimming speed is maximised (approximately 13 body lengths per second). The experimental results were compared to a theoretical model based on three beads, one of which having a fixed magnetic moment and the other two non-magnetic, connected via elastic filaments. The model shows sufficient complexity to break time symmetry and create a net motion, giving good agreement with experiment. Portable point-of-care systems have the potential to revolutionise medical diagnostics. Such systems require active pumps with low power (USB powered devices) external triggers. Due to the wireless and localisation of magnetic fields could possibly allow these portable point-of-care devices to come to life. The main focus of this investigation was to create fluid pump systems comprising from the previously investigated two-ferromagnetic particle swimmer and the torque driven ferromagnetic swimmer. Building on the fact that if a system can generate a net motion it would also be able to create a net flow. Utilising the geometry of the systems, it has been demonstrated that a swimmer-based system can become a fluid pump by restricting the translational motion. The flow structure generated by a pinned swimmer in different scenarios, such as unrestricted flow around it as well as flow generated in straight, cross-shaped, Y-shaped and circular channels were investigated. This investigation demonstrated the feasibility of incorporating the device into a channel and its capability of acting as a pump, valve and flow splitter. As well as a single pump system, networks of the previously mentioned pump systems were fabricated and experimentally investigated. The purpose of this investigation was to utilise the behaviour of the collective motion. Such networks could also be attached to the walls or top of the channel to create a less invasive system compared to pump based within the channel system. The final investigation involved creating collective motion systems which could mimic the beating of cilia - known as a metachronal wave. Two methods were used to create an analogous behaviour. The first was using arrays of identical magnetic rotors, which under the influence of an external magnetic field created two main rotational patterns. The rotational patterns were shown to be controllable producing useful flow fields at low Reynolds numbers. The second system relied on the magnetic components having different fixed magnetisation to create a phase lag between oscillations. The magnetic components were investigated within a channel and the separation between the components was shown to be a key parameter for controlling the induced flow. In both cases, a simple model was produced to help understand the behaviour. Finally, a selection of preliminary investigations into possible applications were conducted experimentally. These investigations included, measuring the effective surface viscosity of lipid monolayers, created cell growth microchannels, as well as systems which could be used for blood plasma separation. The properties of lipid monolayers vary with the surface density, resulting on distinct phase transitions. Slight differences in the molecular lattice are often accompanied by significant changes in the surface viscosity and elasticity. The idea was to use a swimmer as a reporter of the monolayer viscosity, resulting in a less invasive method compared to current techniques to monitor monolayer viscosity, for example torsion pendulums and channel viscometers. The reported effective surface viscosity closely matched the typical Langmuir trough measurements (with a systematic shift of approximately 17 Ų/molecule). The blood plasma separation preliminary work shows the previously investigated two-ferromagnetic particle swimmer mixing a typical volume (100 μm) blood sample with a buffer solution in 21 seconds. The system was also able to create locations with a high population of red blood cells. This resulted in a separation between the blood plasma and red blood cells. Two other preliminary results of future investigations were presented; the collective motion of free swimmers, and the fabrication of ribbon-like structures with fixed magnetic moment patterns.