Manipulation of microparticles using acoustics and Brownian motion

Lab-on-a-chip systems promise to revolutionise laboratory processes in terms of reduced volumes of analytes and reagents, reduced processing time, higher sensitivity, and point of care diagnosis. There are however a number of technical obstacles to be overcome before this promise can be realised. One of these challenges is the ability to concentrate particles within the fluid they are suspended in, or to separate them from each other. In this thesis the separation of particles is studied through the use of numerical modelling, with a number of these studies complementing the experimental work of others. Broadly these can be categorised into concentration of particles within the fluid they are suspended in, separation of particle populations from each other in open systems, and the exploitation of Brownian motion to sort nanometre scale ellipsoidal particles. Within a closed system consisting of a fluid enclosed in a glass capillary, it was found that four equispaced piezoelectric transducers provided the most suitable force field in which particles would be trapped along the axis of the tube. This finite element study considered a range of configurations with one to four transducers and was modelled in two dimensions. The modelling approach was extended to three dimensions to explain the behaviour observed in experiments with a single transducer bonded to a glass capillary. Three-dimensional modelling was required here in order to show how resonance modes can produce particle clump separation of greater than half an acoustic wavelength. Notably all particle clumps were formed away from the transducer allowing visual access around the entire circumference of the capillary. Furthermore a single trapping location existed in the presence of a 0.833 uL/s flow, again away from the transducer. The existence of the single trapping region in the presence of flow was due to the asymmetric placement of the transducer. The dominant force to elicit the behaviour in both of these systems was the acoustic radiation force. The open systems detailed in this thesis consisted of a rectangular chamber or channel with the top surface of the fluid interfacing with air rather than a solid. This broke the symmetry of the systems in one plane and allowed for the creation of particle trapping regions at the fluid-air interface. Two experimental devices were modelled that were able to use the interplay between acoustic radiation forces and streaming induced drag forces to separate different microparticle populations from each other. Particles in the range of microns to tens of microns were found to be able to be sorted in the systems studied. Through the use of modelling it was also shown how the scale of the geometry can affect the relative magnitude of both of these forces; acoustic radiation forces decrease whilst streaming induced drag forces increase. This knowledge can be used to design specific devices for a given pair of microparticles. Finally, a technique for sorting nanometre sized rod-like particles is outlined and demonstrated through the use of numerical modelling. This method utilises a two-stage separation process exploiting the anisotropic Brownian motion of a rod. Simulations of the sorting technique were performed using rods with diameters and lengths ranging from 72 to 168 nm, and 528 to 969.6 nm respectively, achieving 80% population separation within two hours. The particle sizes considered were relatively smaller than in previous studies. The work documented here provides the basis for construction of a physical sorting device.