The use of membranes with a single pore to measure particles as they pass through is a promising method to allow the rapid determination of properties of individual particles in solution. However, this will involve an understanding and control of the forces that these particles undergo while in the pore. In this work we use single cylindrical pores in PET membranes to test and compare the translocations of deformable hydrogel particles and rigid polystyrene spheres. We find that the hydrogel particles cause a characteristic initial increase in current, pass by electroosmosis, and can also undergo dehydration and deformation in pores of smaller diameter. We also show that the force on a particle by an applied hydrostatic pressure gradient can be sufficiently balanced with a transmembrane voltage that is dynamically triggered when the particle enters the pore, and the subsequent fluctuations in the current from the particle diffusing within the pore can be observed for tens of seconds. Finally, we show that repeated translocations of a single particle can be performed using a set rising edge or falling edge triggered voltage wave, or an even greater number of repetitions can be achieved by subtracting in real time the capacitance effects based on a previous test pulse.

Biological nanopores are an essential element to the success of the lipid bilayer that makes life possible on the cellular level. Protein based pores are responsible for the transport of specific ions and molecules across the bilayer and can accomplish their jobs with incredible variety. Some pores, such as those from bacteria have relatively simple structures and allow for many types of transport. Others, like those found in neurons have very high specificity as well as the ability to respond to different cues in the surrounding environment. The vast number of biological ion channels and their complex structures make it difficult to study and understand the nanofluidic phenomena they exhibit. Inspired by the pores seen in nature, scientists have begun to create their own synthetic versions which are usually simpler in structure. These man made pores have been fabricated through a variety of techniques and materials and provide well controlled systems capable of isolating and testing specific transport phenomena. Additionally, these pores provide the building blocks with which more complicated and truly biomimetic pores can be made.

In this thesis, three very different types of nanopores will be examined. First, an alpha hemolysin pore will be isolated and the physics governing transport time will be studied. The diameter and charge of single stranded DNA molecule is modified before being electrophoretically driven through the pores inner constriction. These translocation events are monitored via current measurements across the pore and a biological model is formulated to explain the behavior. This work has the potential to improve DNA sequencing technologies which are dependent on slowing down DNA translocation speed for accurate reads. Next, a synthetic conical nanopore etched in a polymer film will be used as the foundation for a biomimetic nanopore gate. Through the attachment of ssDNA to the pore walls, a pH and voltage responsive nanochannel is created. This is done through the careful selection of the DNA sequence to contain protonatable nucleotides which enable the formation of a transport blocking mesh. Synthetic gated channels that can respond to multiple stimuli as well as have a robust and reversible closing mechanism have been historically difficult to design and are essential to advancing the nanopore field. Finally, a different type of synthetic nanochannel is made from the inner volume of a carbon nanotube. We present a platform that allows for the isolation and study of a single nanotube and use it to examine the underlying transport properties. Carbon nanotube based nanochannels show several unique behaviors which the current literature has not been able to completely explain. Several questions such as the magnitude and origin of enhanced flow and even what carries the current remain open, motivating additional studies, such as ours, which aim to help provide much needed answers.

This thesis describes a novel approach for probing ionic conductivity of silica ionogel electrolytes on the meso- and nanoscale. These properties are important to know for the development of solid-state energy-storage devices that implement nanostructures. Silica ionogels are of specific interest because they maintain a fluidic state on the nanoscale, allowing for a high conductivity of liquid electrolytes, but on the macroscale, they maintain the stability benefits of solid electrolytes. Using single mesopores in polymer films to restrict the geometry of the ionogel to the desired size, we were able to measure the conductivity using two methods. The first method uses liquid electrolytes and makes measurements similar to how the conductivity in liquid electrolytes is found, the other method probes the solid electrolyte in a way more representative to that of energy storage devices. We concluded that the electrochemical properties of ionogel are not significantly affected by constriction to the mesoscale, meaning they can be useful in the implementation of mesoscale solid-state energy-storage devices.

Nanopores and micropores are of great interest as tools in biomimetics and biotechnology. Pores used in this study were prepared by the track-etching technique in polymer foils of polyethylene terephthalate and polycarbonate. Opening diameter of these pores can be tuned from a few nanometers up to tens of micrometers. Nanopores in polyethylene terephthalate were used as a model system for understanding hydrophobic interactions at the nanoscale. Walls of these pores were decorated with alkyl chains of varying length between three and ten carbon atoms. We show that hydrophobic interactions are not only tuned by the pore opening diameter, but also type of chemical modification and electrolyte ionic strength. More concentrated solutions favored wetting of the pores so that finite ion current could be recorded. We also used micrometer sized pores in polyethylene terephthalate and polycarbonate pores as a tool for detecting microspheres and biological cells. We focused on the influence of local diameter undulations of the pores on the object transport. We found that rough forms feature local pressure gradients, which can be used to probe mechanical properties of translocating objects. Potential application of such pores for detecting circulating tumor cells is discussed.

In this dissertation, methods for characterizing cells based on their mechanical phenotypes are described. A novel microfluidic channel design is presented and data are gathered as cells pass through undulations in the channel. Deep learning methods are applied to the data in a new approach for classifying cells solely based on their mechanical properties. First, in a supervised deep learning approach, a highly interpretable random forest was created and trained to extract the most influential features for cell classification. Feature attributions of the random forest were uncovered using Shapley values. Analysis of the most influential features revealed by the Shapley values highlighted the importance of temporal features, such as the change in aspect ratio over time, in classifying cells. This led to the development of a powerful convolutional recurrent neural network, which dramatically improved classification accuracy to more than 90% when using five-fold cross-validation. Next, an unsupervised deep learning approach for cell classification was explored for problems where classes of cells are unknown a priori. Unsupervised clustering was first tested using manually extracted features and traditional clustering algorithms. However, performance was significantly improved with the development of a variational autoencoder (VAE), which extracted higher-dimensional features. The encoder for the VAE was turned into a classifier using a clustering loss function. This trained network exhibited an accuracy of up to 80% when thresholding the top ~10% of the predictions.

Studying ion transport in nanoporous materials is crucial to a wide range of energy and environmental technologies, including ion-selective membranes, drug delivery systems, and supercapacitors. Many of these applications rely on non-aqueous solvents or non-traditional interfaces, such as hydrophobic surfaces. However, classical descriptions of solid-liquid interfaces fail to adequately describe ion and electrical potential distributions in these non-traditional systems. This dissertation examines what consequences occur when these non-traditional solid-liquid interfaces are placed under nanoconfinement. Specifically, ion transport measurements were performed in a variety of nanodevices, including nanopipettes and nanopores, which acted as model systems for probing non-traditional solid-liquid interfaces. My first project details the breakdown in classical electrical double-layer models when moving from water to an organic solvent. This project demonstrates that in a model system, acetonitrile at a silica interface, the solvent molecules organize in a well-defined lipid-bilayer-like structure. This solvent organization in turn dictates the ionic distribution at the interface, which stands in stark contrast to traditional models, where ion distribution is determined by native surface charge. A combined approach of ion-transport measurements in nanopipettes, surface-selective spectroscopy, and molecular dynamics simulations were used to probe the acetonitrile-silica interface in a variety of salts, including LiClO4, NaClO4, LiBF4, and LiPF6. These findings emphasize the importance of including solvent molecules and ions explicitly in descriptions of solid/liquid interfaces. My second and third project examine hydrophobic interfaces in nanopores. Single hydrophobic nanopores are ideal model systems for studying nanoconstricted hydrophobic surfaces and wetting/dewetting transitions at the nanoscale. In these projects, ion current measurements in single silicon nitride nanopores containing a hydrophobic-hydrophilic junction were performed in a variety of salt types and concentrations. The results show that transport properties in these devices are highly dependent on the size, hydration strength, and concentration of the solvated ions. Large, polarizable anions, such as bromide and iodide, facilitate pore wetting, with an unusual dependence on electrolyte concentration – higher concentrated solutions more easily wet the devices, in contrast to bulk surface tension trends. Experimental results were supplemented with molecular dynamic simulations that revealed key characteristics of the asymmetric nanopore system. These results are essential for designing nanoporous systems that are selective for ions of the same charge as well as understanding the fundamental role of ion hydration on the properties of solid-liquid interfaces.

Traditionally, phenotyping cells has been approached through the biological lens of looking at proteins, DNA, RNA, or other molecular descriptions of cells. We present the use of physical properties of cells that emerge as as a consequence of these molecular properties, as a means to mechanically phenotype single cells. In this work, we phenotype two cell types, mammalian and bacterial cells, through separate approaches. In the first study, we create a high-throughput single cell microfluidic stretcher, where we track non-linear dynamics of deforming cells at a high spatial and temporal resolution. We apply image and sequence-based deep learning models for feature extraction that are trained using time sequences of cell shapes, and we found to achieve high classification accuracy based on cytoskeletal properties alone. The best model classified sub-populations with an accuracy over 90%, significantly higher than the 75% we achieved with traditional methods. This increase in accuracy corresponds to a five fold increase in potential enrichment of a sample for a target population. In the second study, we utilize an electrical impedance based approach, called resistive-pulse sensing, to classify bacteria cells based upon their size, shape, and surface potentials. In this method, as individual bacteria cells transit the polymer pore, we observe a changes in electrical impedance proportional to the particle volume, and differences in transit time that is proportional to the surface potential. These features are used to classify three bacteria populations which differ in shape at a high accuracy through unsupervised clustering. We compare this approach to a shape based approach, upon which the resistive-pulse signal is averaged by bacteria of different lengths as they travel along an undulating micropore. This work establishes the application of sequence-based machine and deep learning models to dynamic deformability cytometry and resistive-pulse sensing of bacteria.

Meso- and nano- pores of well-defined geometries and surface charge characteristics are often employed as model systems to probe the electrochemical properties of solid-liquid interfaces. The high surface area to volume ratio and tunable surface charge of these porous systems gives them fascinating ion transport properties that dominate by controlling the solid-liquid interface. These unique transport properties also affect many processes such as the performance of membrane separation (desalination) systems. In this doctoral dissertation, we report experiments designed to facilitate the systematic studying of ionic transport and solid-liquid interfaces using single polyethylene terephthalate (PET) pores in two separate projects:

The first using cylindrical pores, ~ 1 μm in diameter (and 12 μm in length), and the resistive-pulse technique, to investigate the electroosmotic transport of unfunctionalized polystyrene particles through the PET pore in organic, propylene carbonate solutions of LiClO4. The direction of electroosmotic particle translocation informs us about the surface potential of the pore walls and their duration, a measure of the particles’ velocity and the pore walls’ zeta potential. Our experiments show that the carboxlyated, as prepared surface of PET pore(s) become positively charged in LiClO4 solutions of propylene carbonate, even though in aqueous media, the same pore(s) are negatively charged. These findings were also verified by measuring current-voltage curves in a propylene carbonate, LiClO4 concentration gradient. The electroosmotic velocity of the particles also revealed that the positive zeta potential of the pores in propylene carbonate is significantly higher than the negative zeta potential in water.

Additionally, conically shaped pores, 4-25 nm (tip), 700-1100 nm (base) diameter, KCl and bulky chromium, trivalent (Cr3+) cations, were used to probe the switching of the PET surface potential through charge inversion. Briefly, charge inversion occurs when a (negatively charged) pore is placed in contact with oppositely (positively) charged multivalent ions at a sufficiently high (≥ 2) Z and bulk electrolyte concentration. At these experimental conditions, the cations attract, over screen and invert the sign of surface potential here, from negative to positive. The origin of charge inversion is often described through the concept of a strongly correlated liquid that is formed by the multivalent ions in contact with the charged surface. When the surface potential of the pore is inverted, the observable current-voltage curve changes as well and can be experimentally probed. In this work, it was successfully demonstrated through recorded current-voltage curves and ion current signal measurements in time that when the bulk Cr3+ concentration reached 1 mM, the walls of the conically shaped nanopore became positively charged. Also, pores whose tip diameter is less than or equal to 10 nm exhibited ion current fluctuations in time and negative incremental resistance(s). This finding suggests that the pores become partially sterically occluded with the bulky ions which in sub-5 nm pores has major implications on charge inversion and the pore ionic conductance.

Polymer nanopores offer themselves as excellent test beds for study of phenomena that occur on the nano - scale, such as Debye layer formation, surface charge modulation, current saturation, and rectification. Studying ions interactions within the Debye layer, for example, is not possible on the micro - scale, where the pore diameter can be 100 times the size of the zone where interactions of interest occur. However, in our nanopores with an opening diameter less than 10 nm, a slight change of the Debye length can lead to drastic changes of the recorded ion current.

Here we present our nanopores' use as a tool to study geometrical and electrochemical properties of porous manganese oxide. There is great value in studying nano - scale properties of this material because of its importance in lithium ion batteries and newly developed nano - architectures within supercapacitors. We electrodeposited manganese oxide wires into our cylindrical nanopores, filling them completely. In this use, nanopores became a template to probe properties of the embedded material such as surface charge, ion selectivity, and porosity. This information was then reported to the Energy Frontier Research Center (EFRC) collaboration, so that other groups can incorporate these recently discovered characteristics into future their nano - architecture design.

Additionally, we constructed conical nanopores to study interactions between the surface charges found on the walls and alkali metal ions. In particular we looked at lithium, as it is the electrochemically active ion during charge cycling in EFRC energy storage devices. We attempted to reveal lithium ion's affinity to bind to surface charges. We found this binding led to lowering of the effective surface charge of the pore walls, while also decreasing lithium's ability to move through channels or voids that have charged walls. In connection to manganese oxide, a porous, charged material with voids, information on lithium's interaction with these charges is paramount.