The behavior of confined particles in microchannels at moderate Reynolds number has received much attention in recent years and has developed into a new area of research named “inertial microfluidics". This interest has been motivated by the development of high-throughput tools for the manipulation of bioparticles as a precursor for bio-analytic assays. However, a crucial first step towards developing these tools is understanding how particles are transported and localized in confined channels. Here I discuss how the interplay between axial and lateral flow in both a porous and curved microchannel results in non-trivial lateral migration and focusing of finite sized particles. To understand this behavior, I numerically explore this interplay by computing the lateral forces on a neutrally buoyant spherical particle that is subject to both inertial and secondary forces over a range of experimentally relevant particle sizes and channel Reynolds numbers. Interestingly, the lateral forces on the particles in both cases are well represented across a wide range of flow configurations using a simple perturbation based model. The representation of forces in this manner significantly reduces the complexity and time required to predict the migration of inertial particles in microfluidic channels. Finally, I experimentally validate this model and demonstrate how these moderate Reynolds number flows can be used to selectively enrich rare cells in a heterogeneous suspension.

Plasmonic nanoparticles are used in a wide variety of applications over a broad array of fields including medicine, energy, and environmental chemistry. The continued successful development of this class of materials requires the accurate characterization of nanoparticle stability for a variety of solution-based conditions. Although a wide array of methods exist, there is an absence of a unified, quantitative means for complete nanoparticle characterization. This work focuses on the challenges inherent with current methods through a comparative analysis of the current gold standard characterization methods. I propose using capillary electrophoresis and micro-capillary electrophoresis as powerful tools for better quantifying the inherent polydispersity and differences in surface functionalization within a nanoparticle sample. I present the Particle Instability Parameter (PIP) as a robust, quantitative, and generalizable characterization technique based upon UV-Vis absorbance spectroscopy to characterize colloidal instability. I validate PIP performance with both traditional and alternative characterization methods by measuring gold nanorod instability in response to different salt (NaCl) concentrations and as a function of solution pH, salt, and buffer type. I contextualize these methods within the literature on gold nanoparticle characterization to establish a standardized methodology for nanoparticle analysis. Finally, I present a concept for an integrated biodiagnostics platform using gold nanorods based upon an integrated microfluidic microspectrophotometry system for the detection of pathogens.

The electrical double layer (EDL) nano-structure at the interface between electrolytes and charged surfaces dominates the performance of a myriad of electrokinetic and electrochemical processes. A complete understanding of the EDL nano-structure allows for a predictive tool for various systems such as supercapacitors, desalination, and nano-particle manipulation. My work involves developing theoretical models to elucidate the nano-structure of the EDL and the consequent effects on fluid flow and species transport in such systems. These include models explaining dispersion of ions in channels with thick EDLs, surface-charge-based ion conductivity changes, nanofluidic-based DNA hybridization, nanofluidic isotachophoresis, charge inversion due to large ions, and nanofluidic systems with heterogeneous surface charges. Collectively, these studies have enriched our understanding of complex electrokinetic nanochannel transport.

First, I describe a model for the EDL in nanofluidic channels, showing experimentally validated theoretical regimes where dispersion and/or significant EDL size might affect experimental results, as well as methods to account for these effects. Understanding these effects is essential to accurately interpret experiments as well as design of future experiments and subsequent applications. This model can further explain other micro- and nanoscale electrokinetic transport physics. For example, 1) this theory can explain nanochannel conductivity changes due to changes in surface charge, 2) accounting for reaction terms, it can accurately model non-equilibrium DNA hybridization as well as the effect of nano-confinement on such hybridization in electrokinetic capillary electrophoresis-based systems, 3) it can predict an isotachophoretic-like standing front in nanochannels with surface-charge-inverting complex ionic species that induce fluid flow reversal, and 4) it can describe behavior with heterogeneous surface charge.

To explain the behavior of nanofluidic systems with heterogeneous surface charge and complex ionic species, I refined the model by accounting for hard-sphere ion size and more complex near-field screening effects using classical Density Functional Theory. I conducted a theoretical study to explore heterogeneous surface charge in nanochannels with embedded, addressable electrodes that allow us to fully probe EDL structure. I developed a more complete EDL model and performed a systematic theoretical study of EDL nano-structure by varying ion diameter, valence, and concentration, as well as surface charge in order to elucidate EDL nano-structure, fluid flow, and species transport in nanochannels. Thus far we have preliminary model validation using custom-fabricated nanochannels with complex ions, and further experiments will both interpret nanochannel physics through theory as well as improve the model via experimental feedback, overall enabling a more complete predictive theory for future experimental and application design.

Molecules that self-assemble to form thin coatings on surfaces are of great interest to many industries (e.g., biomedical, marine, and solar) for imparting hydrophobic, conductive, and/or anti-fouling properties. For example, in capillary electrophoresis (CE), inner surfaces of a capillary are coated to control electroosmotic flow (EOF), prevent adsorption, and increase solubility of proteins; this coating reverses the direction of EOF relative to electrophoresis (EP), thereby enhancing separation efficiency and resolution. The rapid degradation of these coatings, however, causes band broadening and variable migration times as the desorbing species changes the EOF velocity and interferes with the detector. Unfortunately, studies to understand desorption in these conditions show different kinetic behaviors and employ low-resolution analysis techniques. Therefore, techniques with greater temporal resolution and control of solution and transport properties are needed to elucidate degradation mechanisms in CE and other applications.

This work presents a high frequency, in situ platform for extracting surface charge (i.e., zeta potential) within a silica microcapillary during surface coating or degradation. A continuous platform was adapted from the current monitoring method, which extracts EOF velocity by measuring the time for fluid to traverse a channel of known length in an electric field; the traversal endpoint is measuring by a linear change in current and subsequent steady state as a solution of different conductivity fills the channel. Zeta potential is calculated using the Helmholtz–Smoluchowski equation and the EOF velocity. Automated zeta potential analysis (AZA) detects the traversal endpoint using the first-derivative, then alternates the electric field polarity for consecutive measurements.

This novel automated approach for monitoring zeta potential was first applied to study aminosilane coating formation and stability for different monomer types and solution conditions. The density of cationic coatings was inferred by an increase in zeta potential relative to the bare surface value; coating stability was monitored by the relative decay in zeta potential after exposure to conditions without the monomer. First, applying AZA to study aminosilanes, we observed higher densities for trimethoxy vs. monomethoxy silanes and faster coating kinetics for trimethoxy vs. triethoxy silanes. Upon exposure, we observed faster decay for higher-pH solutions and shorter-alkyl-length monomers. Coatings deposited in aqueous vs. anhydrous conditions exhibited a lower density and greater stability, indicating that aqueous conditions promote more lateral crosslinking.

We also applied AZA to study adsorption and desorption of CTAB, a cationic surfactant, in conditions relevant to CE (e.g., capillary diameter, EOF/EP, and pressure). In contrast to previous studies, our kinetics were multiphasic with an intermediate "stagnant regime” at distinct zeta potentials in adsorption or desorption. These stagnant regimes occurred at positive zeta potentials where the EOF was equal and opposite to EP, inducing a near-zero net transport of CTA+ (EP+EOF). The kinetics depended strongly on the capillary surface-area-to-volume ratio and the EP mobility of CTA+, which changed with concentration. We confirmed these trends by recasting the zeta potential kinetics in terms of net CTA+ transport volume divided by surface area; this normalization collapsed and linearized the responses for a range of diameters. We showed that varying the voltage algorithm and applying pressure could prolong or eliminate the stagnant regime; new AZA algorithms were developed to further control the transport-limited kinetics by maintaining a fixed net transport of CTA+. With higher resolution and control of transport during coating formation and degradation, this versatile platform enables the discovery of optimal coating chemistries for an array of applications.

DNA-templated silver nanoclusters (DNA-AgNCs) are fluorescent molecules containing few-atom clusters of silver stabilized by a short strand of DNA. They fluoresce at wavelengths in the visible spectrum with emission colors tuned by varying the sequence of the DNA used for their synthesis. The combination of their small size, tunable spectra and biocompatibility opens exciting, new possibilities for their use in chemical sensing, biological labeling and imaging, genetic mutation detection and the development of self-assembled DNA nanotechnologies.

A detailed understanding of the structure of DNA-AgNCs is vital for precisely engineering their properties for future applications. Specifically, changes in the arrangement and composition of bases surrounding a AgNC can strongly influence its fluorescence. Although this sensitivity has already been leveraged for the development of novel chemical sensors, a better understanding of the mechanism is needed to improve performance and broaden the applicability of DNA stabilized AgNCs. In this work, we use high-resolution microfluidic capillary electrophoresis to show that, although AgNC depend on single-stranded DNA for their stabilization, changes made to bases in the double-stranded stem region of a DNA hairpin can perturb their structure, leading to differences in fluorescence emission.

We also explore ways in which AgNC can be robust to DNA sequence changes. We document one DNA-AgNC in which the DNA adopts different conformations, or shapes, that yield clusters with the same emission color. We also show that poly-thymidine regions, which are know to bind silver poorly,can act as convenient handles to adjust the electrophoretic mobility of DNA-AgNCs without affecting their fluorescence.

Using poly-thymidine appendages, we demonstrate a way to tune the electrophoretic mobility of a AgNC-based DNA probe for use in a microfluidic assay. These label-free probes are composed of a DNA hairpin that generates a fluorescent silver nanocluster only after binding to a specific target DNA sequence. By tuning the mobility of probes designed to bind different targets, we demonstrate a rapid microfluidic separation assay for the multiplexed fluorescent detection of nucleic acid targets for Hepatitis A, B and C.

The probe design initially used for these studies suffered from several shortcomings. Most significantly, the probe DNA failed to generate a fluorescent AgNC for some binding domain sequences. To overcome this, as well as provide added functionality, we engineer a new DNA-AgNC based sensor that supports ratiometric fluorescence measurements for the sensitive, specific and low-cost detection of DNA. Probes based on our new design generate a green emitting AgNC in its hairpin state, and a red emitting AgNC after binding their target. The ratiometric fluorescence provides a stable signal and rapid quantification of DNA concentration regardless of the choice of target, a dramatic improvement over similar turn-on fluorescent probes.