3D Phononic-Fluidic Systems for Liquid Mixture Analysis and Control

This project aimed at realizing and optimizing phononic-fluidic sensor systems, by combining acoustically significant microstructures with fluidic elements. This combination paves the way for the acoustic analysis of a variety of fluids, from simple liquids like water and alcohols, clean mixtures and solutions to complex liquids like cell cultures or blood. Research on acoustic metamaterials and phononic crystals is the theoretical foundation of these devices. Modern additive manufacturing forms the enabling technological basis to realize complex three-dimensional geometries. Phononic crystals, acoustic equivalent of photonic crystals from optics, offer unique band structures to manipulate acoustic wave propagation. Additive fabrication was used to realize 3D phononic crystals with band gap characteristics beyond the capabilities of corresponding 2D designs. Adding a fluidic volume into a phononic lattice as defect was the first device concept explored in this work. A phononic-fluidic cavity defect acts as an acoustic resonator that combines the high sensitivity of resonant sensors with the ability of ultrasonic sensors to probe volumetric properties of fluids. The cavity defect resonance is designed to fall into a phononic band gap and yield a highly sensitive and unperturbed resonance peak dependent on the physical properties of the liquid analyte, such as speed of sound, density, viscosity, and subsequently concentration of mixtures. We successfully fabricated phononic-fluidic sensors using complex 3D geometries and demonstrated the functionality and advantages by measuring the concentration of aqueous sodium chloride and glucose mixtures, as well as later for other liquids and mixtures. Our relatively simple sensor elements achieved remarkable sensitivity and resolution. An important conclusion is the necessity to model such sensors taking all real-world constraints into account. For numerical studies precise, frequency-dependent material data including viscous and solid losses is critical. Likewise, the finite size of a real sensor element introduces additional constraints compared to 1D or 2D approximations, potentially destroying any sensor effect found only in idealized cases. During the project, we were also able to explore the potential of numerical optimization methods. First results highlight the possibility of optimizing 3D designs to achieve a significantly better sensor performance and generate completely novel design ideas. We believe that this is one potential path for further developing this exciting research topic.