Nanofluidic devices for nanoparticle imaging

Exosomes are 30-150 nm extracellular vesicles that have gained significant interest over recent years, as advances in cell imaging have revealed their role as cancer biomarkers. Their natural abundance in bodily fluids makes them ideal candidates for early, non-invasive diagnostics. However, their purification and characterization remain significant challenges. Exosomes occur at low concentrations relative to other circulating components-such as proteins, lipoproteins, and cellular debris-and are found in complex biofluids, demanding sophisticated isolation strategies. Existing purification techniques, including size-based separation and immunoaffinity capture, often lack selectivity, require costly reagents, or involve lengthy processing times. Characterization is equally challenging due to the heterogeneity of exosome morphology, molecular composition, and cargo. Current analytical methods frequently suffer from poor specificity, prolonged processing, or sample degradation, and conventional workflows typically separate isolation and analysis, which can compromise sample integrity.

This thesis presents a nanofluidic platform that leverages the interplay between diffusiophoretic particle transport and diffusioosmotic fluid flow to isolate and characterize vesicles based on their zeta potential. Rasmussen et al. demonstrated the feasibility of this approach, and this work extends it by combining selective isolation with the trapping mechanism. Specifically, the platform exploits the open-ended geometry of the trap to allow certain vesicle populations to escape while others remain confined.

I first introduce a novel theoretical framework in which vesicles can escape the trap if their trapping positions are near its exit. I derive analytical expressions for the vesicle distribution and current. From this, I use numerical simulations to determine optimal experimental parameters (trap geometry, wall coating and salt gradient) that shift the distribution toward the trap’s outlet. I design and fabricate an improved nanofluidic platform to expand experimental flexibility.

I develop new phospholipid formulations for wall coating and adapt the experimental protocol accordingly. To demonstrate controlled release, I use synthetic liposome populations with different fluorescent dyes. My results show that two populations differing by 7 mV in zeta potential exit the trap at distinct rates, achieving a 1.5-fold enrichment of the low-zeta-potential population within the trap compared to the input.

To demonstrate the platform’s robustness in the presence of proteins, I replace the lipid wall coating with the amphiphilic polymer Pluronic F127. The modified system maintains its selective trapping functionality, confirming the coating’s stability and the technique’s applicability in complex media.

Taken together, this work establishes a novel nanofluidic enrichment method for vesicle isolation and characterization using diffusiophoresis and diffusioosmosis. By adjusting vesicle distribution positions and subsequent currents within the trap, the method enables precise enrichment and depletion based solely on diameter and zeta potential. This enrichment capability extends beyond exosomes to other vesicle types, advancing next-generation cancer diagnostic technologies.