Mass spectrometry (MS) analysis of the cell secretome could help identify promising candidates for directed evolution, but it requires separating the cell from its secretions to keep it alive. In this project, we present a droplet microfluidic method that uses acoustic forces to non-invasively manipulate and separate particles or cells from the surrounding supernatant within nanoliter droplets. The approach achieves up to 100% separation efficiency at 114 droplets/min for beads and yeast cells, and is versatile enough to handle mammalian cells and bacteria, providing a robust strategy for single-cell secretome sampling.

Genetically modified bacteria are powerful tools for producing tailored proteins, but reprogramming often requires cumbersome medium exchange steps. We developed an acoustofluidic device that enables non-invasive trapping and efficient medium exchange of E. coli, overcoming challenges posed by acoustic streaming. Operating at 10 μL/min, the device achieves 47 ± 3% cell recovery, while maintaining high electrocompetence with transformation efficiencies of 8 × 10⁵ CFU/μg DNA. This low-volume alternative to centrifugation simplifies and miniaturises a critical step in synthetic biology, advancing automation of microbiological and molecular engineering workflows.

In this project, we set out to improve the resolution of metal 3D printing by Direct Metal Laser Sintering, where the main limitations arise from the laser focus spot size and the metal powder particle dimensions. To address this, we developed an acoustofluidic approach that enables two-dimensional focusing of metal microparticles in acoustically excited round glass capillaries, offering a low-cost alternative to cleanroom-fabricated devices. Using bulk acoustic waves, we achieved robust focusing of 1 µm copper particles into a narrow line (60.8 ± 7.0 µm width, 45.2 ± 9.3 µm height) with a 90-fold local concentration increase, and even manipulated 1 µm polystyrene particles typically hindered by acoustic streaming. Numerical analysis revealed that unique streaming transitions lower the critical particle radius compared to rectangular channels, enabling applications such as particle ejection through 25 µm openings for processes that usually suffer from clogging, abrasion, or limited resolution.

Deep reactive ion etching (DRIE) using the Bosch process enables high-aspect-ratio silicon microstructures for MEMS and microfluidics but is limited by etch lag and complex mask requirements. We present an optimized two-step Bosch process that reduces etch lag to below 1.5%, significantly improving etch uniformity. In addition, we demonstrate a three-step process capable of producing stable 6 μm wide structures with depths up to 180 μm, advancing miniaturization for biomedical applications.

Protein aggregation caused by shear and interfacial stress is a major challenge in therapeutic protein and antibody production. We developed a microfluidic device that exposes small protein volumes to uniform shear, enabling us to separate flow effects from interfacial contributions by comparing PMMA and stainless-steel devices. Our results show that interfaces strongly mediate aggregation, with steel producing 3.5-fold fewer particles than PMMA, while surfactants prevent aggregation entirely—highlighting the device as a valuable tool for formulation stability testing in biopharmaceutical development.