Multidimensional particle fractioning using surface acoustic waves

点击翻译按钮获取中文摘要

For many industrial applications, in process engineering and food chemistry as well as pharmaceutical production, particulate systems are of vital importance. In microfluidics, forces acting on particles scale differently due to the square cube law. Here passive and active methods can be applied for fractionation. Passive methods such as inertial flow or flow focusing have the benefit to be very robust and do not require external energy. However, active methods applying external fields (such as magnetic, electric, and acoustic fields) offer a greater flexibility at the expense of more complex systems. A combination of both types opens the possibility to build a flexible and controllable system that can be adapted to the different parameters of varying particle suspensions. Therefore, these systems are well suited to fractionate particle suspensions by different properties. The aim of the research project is to setup a system using surface acoustic waves to manipulate and sort particles due to their size, shape and acoustic contrast (density and compressibility). In order to be able to characterize such systems, a deep neural network should be applied to learn the particle position based on astigmatically deformed particle images. In addition, the network can be trained to extend the applicability of this measurement technique to multi dispersed particles and to detect the corresponding size and shape of the particles. Finally, microfluidics fractionation systems shall be produced at the technology center (Center of Micro- and Nanotechnologies) at TU Ilmenau and characterized in detail, to draw conclusions on the design of technical relevant systems for the customizable fractioning of suspensions of very fine particles (< 500 nm).