Development and ssage of micro- and nanofluidic devices for nanoparticle trapping, sorting and biosensing

Microfluidics has revolutionized life sciences by introducing the tools to perform complex scientific studies in a simpler yet robust and reliable way. Miniaturization of bench-top processing tools using micro- and nanofluidic devices enables handling biological samples in a physiologically relevant environment to execute complex studies that were not possible before. Organ on a chip, lab on a chip, point-of-care diagnosis, biosensing, miniaturized PCR tools, etc., are some of the previously inconceivable examples in a portable device form. Due to the scale of the device dimensions in such microfluidic devices, small volume handling and processing have become noticeably effortless.
Among various applications of micro- and nanofluidic devices, molecular sensing, nanoparticle separation, sorting, trapping, and processing are of significant impact due to their feasibility of implementation in most of the fluidic devices. Single-particle trapping is an effective approach to study the fundamental properties of molecules in their physiological environment. Various active and passive methods exist to execute single-particle studies, such as optical tweezers, magnetic tweezers, dielectrophoretic trapping, hydrodynamic trapping, geometrical trapping, and electrostatic trapping. In the case of active methods, such as optical and magnetic tweezers, precise control of molecular motion is possible at the cost of a complex setup with external force sources. However, high-throughput single-particle trapping and manipulation are not feasible in a way that can be achieved using passive methods such as geometry induced electrostatic (GIE) trapping and geometrical trapping.
This thesis focuses on developing integrated micro and nanofluidic devices for 1) high throughput contact free electrostatic trapping of single nanoparticles and 2) size based nanoparticle separation, sorting, and trapping for biosensing applications. The high-throughput single-particle trapping was achieved by developing fluidic devices utilizing the GIE trapping. A GIE trapping fluidic device comprises nanochannels embedded with nanostructures, such as slits, cylinders, and grids. These nanostructures enable the formation of electrostatic potential traps inside the nanoindentations, forcing negatively charged nano objects to attain a position inside them to minimize their self-energy. In conventional GIE trapping devices, negatively charged molecules, such as DNA, viruses, and gold nanoparticles (Au NPs), can be easily trapped in the electrostatic traps.
This thesis presents the development and fabrication of GIE trapping devices using 1) glass substrate and 2) polydimethylsiloxane (PDMS) polymer. These substrates attain a net negative surface charge density in an aqueous solution (pH > 2) due to the self-dissociation of terminal silanol groups. Therefore, glass and PDMS based fluidic devices are only usable for the confinement of negatively charged nano objects. In this work, the scope of these fluidic devices was extended to the trapping of positively charged nano objects by using surface modification methods for both glass and PDMS based fluidic devices. The surface modification of glass‑based nanofluidic devices was achieved by modifying the inside of the GIE-trapping device by the adsorption of a single layer of polyelectrolyte (poly(ethyleneimine), PEI). The PEI layer modifies the negatively charged glass surface to a positively charged surface and allows for the trapping of positively charged nanoparticles. However, the surface modifying procedure for the glass based GIE trapping device was demanding and required 4 to 5 days. To have an efficient surface modification process, PDMS based GIE trapping devices were introduced.
The introduction of PDMS based fluidic devices for positively charged nano objects has improved the throughput for device fabrication and surface modification. Furthermore, two polyelectrolyte layers (1: poly(ethyleneimine) and 2: poly(styrenesulfonate)) deposition is presented in this work using PDMS based devices to demonstrate the possibility of achieving homogeneously charged surface using multi-polyelectrolyte layers. The efficiency of these devices with surface charge reversal was comparable to native GIE trapping devices, demonstrating the successful and homogeneous surface modification.
The trapping efficiency and device performance of a GIE Trapping device rely on the geometry of the device and the interaction between the charged particle and the device surface. Therefore, extensive optimization of the device geometry is essential to achieve efficient GIE trapping in a fluidic device. In this work, two different approaches, 1) charged particle inclusive simulation and 2) point charge approximation simulation, are presented to optimize the geometrical parameters of a GIE trapping device numerically. To compare numerical results with experimental data, a cylindrical nanopocket design was used to represent a nanotrap to confine a charged gold nanoparticle.
The charged-particle inclusive simulations are demanding, but provide more accurate results for attainable particle stiffness constant using crucial geometrical parameters of the device, size and charge of the particle of interest, and the salt concentration of the solution. Comparatively, point-charge approximation simulations are faster and give appropriate results of particle trapping stiffness constant, residence time, etc. Here, point-charge approximation simulations are used for efficiently identifying the trends of trapping strength of a device based on critical geometrical parameters, i.e., the height of the nanochannel and the nanopocket and the diameter of the nanopocket. The point charge approximation simulations demonstrated that the trapping strength of a particle inside a nanotrap could be enhanced by increasing the trap height or reducing the channel height. Additionally, the trapping strength of a nanotrap can be modified by changing the diameter of the nanopocket; however, reduction or enlargement of the pocket diameter from the optimum diameter reduces the trapping strength of the nanotrap. For effective GIE trapping, it is important to use a solution with low ionic or salt concentration (< 0.5mM for trapping stiffness constant > 10-4 pN/nm) in order to avoid screening of the electrostatic field from the charged device surface. A detailed comparison of both approaches, numerical calculations, and experimental results are presented, demonstrating their advantages and disadvantages.
While there are many advantages of GIE trapping devices for molecular trapping, one major disadvantage is the reduced functionality of the devices for body fluids that contain high salt concentrations. Due to the high ionic concentration in the body fluids, the electrostatic effect of the charged device surface gets screened, leading to no potential trap for the confinement of charged nano objects. Therefore, a new design of the fluidic device is developed for biosensing applications that can use body fluids to extract the target molecules for molecular sensing. The fluidic device exploited geometrical sieving, deterministic lateral displacement (DLD) arrays, and geometrical trapping for particle separation, sorting, and trapping, respectively. The separation of unwanted macro- and micro-particles was achieved in the separation chamber, followed by the size sorting of target molecule adsorbed nanoparticles and, later, the size based trapping of these nanoparticles in the detection area. The motion of the solution and nanoparticle throughout the device was observed using interferometric scattering detection (iSCAT) microscopy, whereas, for molecular sensing, Raman spectroscopy was used at the detection area to achieve a few pg/ml detection limit. The device has the potential for applications in early multi disease diagnosis for diseases that can be detected using antigen-antibody complex formation on antibody-coated nanoparticles.
The presented GIE trapping devices can be used to achieve high-throughput single-nanoparticle trapping, whereas geometrical particle trapping devices can be used to perform size-selective nanoparticle trapping for molecular sensing. Both methods are effective for studies conducted in an aqueous environment and have the potential to be used in molecular studies, disease diagnosis, biological studies, etc., for research and commercial purposes. Demonstrated device fabrication methods and surface modification procedures allow improved productivity and yield of the GIE trapping devices. The device geometry of a GIE trapping device can be optimized further using the presented numerical calculations. Therefore, the work presented here advances the research in the field of GIE trapping and geometrical trapping and opens up new possibilities for both basic and applied research in several fields, such as biophysics, molecular dynamics, diagnostics, and molecular detection.