Ce travail de thèse s’intéresse aux interactions entre la solution et la pellicule d’altération, appelée gel, qui se forme lors de la lixiviation des verres borosilicatés d’intérêt nucléaire, et plus spécifiquement dans ce travail, sur le verre à six oxydes ISG. Afin de mieux comprendre les mécanismes de formation du gel, ainsi que ses propriétés de passivation, une étude combinant analyses expérimentales et simulations de dynamique moléculaire (DM) est réalisée.Dans un premier temps, l’étude se focalise sur la structure du verre sain dans le but de mieux appréhender les processus de lixiviation. Pour ce faire, des analyses par spectroscopie RMN sont réalisées. Les résultats sont comparés à ceux obtenus par DM afin de valider la simulation. Cette dernière permet alors une exploration complète de l’ordre à courte et moyenne distances. Elle montre en outre une distribution homogène des formateurs de réseau et l’absence de zones enrichies en bore. Ce dernier, que l’on sait fortement soluble, a tendance à fragmenter le réseau silicaté, affectant de ce fait sa réactivité.Les processus de formation du gel sont ensuite étudiés expérimentalement, dans des conditions favorisant la passivation (90 °C, pH 7, solution saturée en silice). Nous mettons en évidence un fort effet des éléments exogènes en solution, notamment les alcalins faiblement hydratés tels que le potassium et le césium. Ces derniers entrainent ainsi une chute marquée de l’altération du verre. Afin de comprendre cet effet, les gels formés en présence de différents alcalins sont caractérisés. Les analyses montrent un départ congruent du bore et du sodium, et une incorporation du potassium et du césium de la solution qui prennent le rôle de compensateur de charge des unités [AlO4]- assuré par le calcium dans les gels formés en milieux sans alcalin ou contenant des alcalins fortement hydraté comme le lithium ou le sodium. Les différents gels présentent des degrés d’hydratation variés, en accord avec les résultats de simulation de DM de la diffusion de l’eau dans une silice nanoporeuse en milieu alcalin. Ces simulations montrent que la combinaison d’effets stériques (taille des alcalins) et physico-chimiques (énergie d’hydratation) entraine une diminution de la quantité d’eau dans les porosités contenant des ions potassium ou césium.Par ailleurs, les différents gels obtenus dans les conditions d’altération précitées sont fortement polymérisés, indiquant une réorganisation du réseau silicaté à la suite du départ des éléments facilement hydrolysables comme le bore. Cette réorganisation a lieu sans hydrolyse complète des tétraèdres de silicium, ce qui invalide dans ce cas le modèle de formation des gels par dissolution congruente/précipitation. La spéciation de l’eau au sein du gel, obtenue en combinant ATG et RMN du proton, permet de plus de déterminer quantitativement la répartition des atomes d’oxygène au sein du gel (atomes d’oxygène pontants, non pontants ou H2O).Ces données sont alors utilisées pour l’interprétation d’expériences de traçage en milieu enrichi en eau marquée en 18O d’échantillons pré-altérés. Ces expériences, menées à différentes températures, mettent en évidence pour la première fois une réorganisation continue du réseau du gel, avec une diminution au cours du temps de l’accessibilité des espèces mobiles (H2O et hydroxyles) au verre sain du fait de la maturation de la porosité au sein du gel. Nous déterminons différentes diffusivités pour l’eau en fonction de la topologie du gel et proposons l’hypothèse que cette atténuation progressive de la réactivité du réseau soit à l’origine du caractère passivant du gel. / This work aims at understanding water interactions with the altered layer, called gel, formed during borosilicate nuclear glass corrosion. Specifically, we focus on the corrosion of the six oxides ISG. To better understand gel formation mechanisms, as well as gel passivating properties, experimental studies are combined to molecular dynamics simulations.First, this study focuses on the characterization of the pristine glass structure. Experimental analysis (NMR) provides some information to validate the simulated structure. As a result, an improved understanding of the pristine glass short- and medium-range orders is obtained. We also observe that network formers are homogeneously distributed, with no area enriched in boron for instance. Boron is known to be highly soluble, and tends to divide the silicate network, which would affect its reactivity.The gel formation is then studied experimentally in conditions favoring the passivating effect of the glass (90 °C, pH 7, silica-saturated solution). A strong effect of exogenous elements in solution, particularly weakly hydrated alkalis such as potassium and cesium, is observed, with a notable decrease of glass corrosion. To better understand this effect, all gels are characterized. A congruent release of boron and sodium is observed, while potassium and cesium are incorporated. They then act as charge compensator for [AlO4]- units. Calcium usually plays this role in gels formed in solutions with no alkali or containing strongly hydrated alkali such as lithium and sodium. The hydration degree differs for the various gels present, as confirmed by MD simulation of water diffusion in nanoporous amorphous silica in presence of alkali. These simulations highlight a combined impact of sterical effects (alkali size) and physicochemical effects (hydration energy) leading to a decrease of water quantity in the nanopores containing potassium and cesium.Moreover, all the gels formed in the above mentioned leaching conditions are highly polymerized, which indicates a reorganization of the network following the leaching of hydrolysable species such as boron. This reorganization happens without complete hydrolysis of silicon atoms, which invalidate the congruent dissolution/precipitation model in this case. Water speciation inside the gel is determined combining TGA and NMR, giving access to quantitative oxygen repartition in the gel (bridging oxygen, non-bridging oxygen or H2O).This data are then used for the interpretation of tracing experiments carried out in H218O rich solution on prealtered samples. The results demonstrate for the first time that the network reorganizes continuously over time, with a decrease of mobile species (H2O and hydroxyls) accessibility due to the maturation of the porosity within the gel. We determine various water diffusivities as a function of the gel topology and propose the hypothesis that this decreasing reactivity of the network is the source of the passivating nature of the gel.

La récupération d'hydrocarbures non conventionnels fait partie des enjeux énergétiques majeurs. Ils ne peuvent être extraits par simple forage car la roche qui les contient, constituée essentiellement de nanopores, présente une très faible perméabilité. A l'échelle macroscopique, c'est-à-dire à l'échelle du bassin, les écoulements de fluides sont décrits par la loi de Darcy qui relie le flux à la perméabilité, au gradient de pression et à la viscosité. La perméabilité d'un matériau peut être mesurée expérimentalement ou théoriquement par homogénéisation à partir de l’hydrodynamique continue. Cependant, lorsque la taille des pores devient comparable à celle des molécules de fluide, une telle description n'est pas satisfaisante. D’une part l’hydrodynamique continue, où la nature du fluide n’intervient qu’à travers la viscosité, ne suffit pas forcément pour décrire l’écoulement. D’autre part les interactions au niveau moléculaire entre le fluide et le solide jouent un rôle important. Cette thèse porte sur le transport de fluides à l'échelle moléculaire et revisite la description traditionnelle qui sert de point de départ pour des écoulements à l'échelle macroscopique, en particulier dans le cas des écoulements multiphasiques. Par des simulations de dynamique moléculaire classique, nous avons étudié l'écoulement de systèmes monophasiques et diphasiques, précisant l’influence de la nature des surfaces, ainsi que de la nature et de la concentration des espèces dissoutes. Nous avons également apporté une contribution méthodologique originale pour le calcul des coefficients de diffusion d'espèces. / Unconventional hydrocarbons recovery is one of the major energy challenges. They cannot be extracted by simple drilling because the rock which contains them, consisting essentially of nanopores, has a very low permeability. On the macroscopic scale of the geological basin, the flows of fluids are described by Darcy's law which connects the flux to the permeability, the pressure gradient and the viscosity. The permeability of a material can be measured experimentally or determined theoretically by homogenization from continuous hydrodynamics. However, when the pore size becomes comparable to that of the fluid molecules, such a description is unsatisfactory. On the one hand continuous hydrodynamics, where the nature of the fluid only enters via the viscosity, is not necessarily sufficient to describe the flow. On the other hand, the interactions at the molecular level between the fluid and the solid play an important role. This thesis deals with the transport of fluids on a molecular scale and revisits the traditional description which serves as a starting point for macroscopic flows, in particular in the case of multiphase flows. Using classical molecular dynamics simulations, we study the flow of one- and two-phase systems, specifying the influence of the nature of the surfaces, as well as the nature and concentration of the dissolved species. We also develop an original methodological contribution to the calculation of the diffusion coefficients of confined species, specifying the effects of the system finite size.

Nanopore sensors are emerging as a promising technology for single molecule analysis and polymer sequencing. Traditionally, measurements are taken by monitoring the ionic current through the nanopore, which gives information (e.g. size, shape, charge) about a molecule of interest while it is in the confined geometry of the nanopore. The dynamics of the molecule before the arrival to the nanopore, such as the capture dynamics, or molecular conformation prior to translocation, as well as clogging mechanisms and features of anomalous translocation events, are not assessed by the electrical measurements alone. To study the whole process of nanopore diffusion, capture and passage it is necessary to complement the electrical signal with another detection mode. Particularly, optical visualization of the molecules as they translocate through the nanopore has great potential. In this Thesis I present the design, construction, optimization and testing of a nanopore--‐based optofluidic instrument, which uses fluorescence microscopy to visualize individual fluorescently stained DNA molecules as they translocate a solid--‐state nanopore, while in parallel record the ionic current signal through the pore. The following challenges were overcome to achieve the integration of the optical and electrical systems: (i) the electrical detection system must account for the physical constrains of a wide field fluorescence microscope, and the optical system should in turn not affect the low--‐noise electrical detection of individual DNA molecules. The design of the instrument included a microfluidic device, so to position the nanopore within the working distance (<170--‐μm) of the microscope objective (Chapter 2). (ii) Electrical noise was optimized to a level that is indistinguishable from a standard (with no optics) nanopore system (Chapter 3). The custom instrument was used to demonstrate: 1) Electrical detection of DNA translocations with a laser light illuminating the nanopore; 2) Optical tracking of DNA capture and translocation dynamics; 3) Synchronization of the optical and electrical signals in preparation for simultaneous detection. In the process of noise optimization, a strong noise coupling between the illumination source and the ionic current was found, characterized and eliminated. Consequently, the noise performance of the custom instrument is the lowest of any other nanopore--‐based optofluidic systems described in the literature to date. This opens up the way to many new and exciting investigations of polymer translocation dynamics through nanoconfined geometries. Lastly, during the development of this custom instrument, a method to localize the fabrication of a nanopore by controlled dielectric breakdown on a membrane, with a focused laser beam, was discovered.

Yes / The next-generation short-read sequencing technologies that generate comprehensive, whole-genome data with single-nucleotide resolution have already advanced tuberculosis diagnosis, treatment, surveillance and source investigation. Their high costs, tedious and lengthy processes, and large equipment remain major hurdles for research use in high tuberculosis burden countries and implementation into routine care. The portable next-generation sequencing devices developed by Oxford Nanopore Technologies (ONT) are attractive alternatives due to their long-read sequence capability, compact low-cost hardware, and continued improvements in accuracy and throughput. A systematic review of the published literature demonstrated limited uptake of ONT sequencing in tuberculosis research and clinical care. Of the 12 eligible articles presenting ONT sequencing data on at least one Mycobacterium tuberculosis sample, four addressed software development for long read ONT sequencing data with potential applications for M. tuberculosis. Only eight studies presented results of ONT sequencing of M. tuberculosis, of which five performed whole-genome and three did targeted sequencing. Based on these findings, we summarize the standard processes, reflect on the current limitations of ONT sequencing technology, and the research needed to overcome the main hurdles. Summary: The low capital cost, portable nature and continued improvement in the performance of ONT sequencing make it an attractive option for sequencing for research and clinical care, but limited data is available on its application in the tuberculosis field. Important research investment is needed to unleash the full potential of ONT sequencing for tuberculosis research and care.

Nanopores have been shown to be a useful analytical tool for single molecule detection. They have been used to study the composition of DNA and other molecules of interest. These pores are usually α-hemolysin which is a toxin from Staphylococcus aureus or more recently nanoscale synthetic solid state pores. Now we are beginning to look at other molecules or proteins by sending them into the nanopores and measuring a characteristic partial current blockade. In this thesis we look at polyethylene glycol (PEG) as it enters and blocks current through a single alpha hemolysin pore. We report the effects of ionic strength, PEG size, and applied voltage on the depth and duration of the current blockades. We also apply autocorrelation analysis on the arrival times of PEG molecules to the pore see if we can identify if the PEG is translocating through the pore or escaping from the same side it enters. This suggests a new approach to current blockade analysis.

A DNA molecule passing through a nanopore in a liner and sequential fashion allows for unprecedented interrogation of the polymer. Adding transverse electrodes that are comparable in size and sensitive to the DNA molecule, can further the attempts to rapidly sequence DNA. Carbon nanotubes are comparable in size and interact strongly with the DNA molecule. This makes them an excellent choice for integration with nanopores. Only the section of the carbon nanotube in immediate proximity to the nanopore should be sensitive to the DNA molecules. Atomic layer deposition of metal-oxides passivates the sections of the carbon nanotube that are not to interact with the DNA molecule. The coating also protects the thin film interconnects leading to the carbon nanotube. Hafnium oxide is superior to aluminum oxide in chemical resistance and electrical insulation but leads to a high failure rate of the carbon nanotube across nanopore devices. Aluminum oxide, combined with gold thin film interconnects to the carbon nanotube, produced the first functioning devices in electrolyte. These devices had concurrently functioning ionic (current across the nanopore) and transverse (current through the carbon nanotube) channels. No concurrent DNA translocation signal was recorded on the ionic and nanotube current traces. Analyzing the translocation events recorded on the ionic channel indicated that double-stranded DNA (dsDNA) passed through the carbon nanotube articulated nanopore an order of magnitude slower than it would have through a comparable unarticulated nanopore. The slower translocation observed is a necessary condition for sequencing. Investigating dsDNA translocation under various experimental conditions led to the discovery of a new interaction between the molecule and small nanopores. A dsDNA molecule is trapped when the electric field near the nanopore attracts and immobilizes a non-end segment of the molecule at the nanopore orifice without inducing folded translocation. In this work, the expression “trapped dsDNA” will exclusively refer to the immobilization of a dsDNA molecule at the orifice of the nanopore. The ionic current through the nanopore decreases when the dsDNA molecule is trapped by the nanopore. By contrast, a translocating dsDNA molecule under the same conditions causes an ionic current increase. Finite element modeling results predict this behavior for the conditions of the experiment. / Physics

The work presented in this thesis is divided loosely into three main areas of interest: development of a novel method of solid-state nanopore fabrication; applications of this method to some of the open problems in the field; and analysis of nanopore data. The first of these occupies the majority of the research presented in this thesis, covering research dedicated to the development and characterization of a novel method of solid-state nanopore fabrication which achieves nanometer scale control over matter using simple and low cost circuitry. Termed controlled breakdown (CBD), this technique is in the process of revolutionizing the field of nanopore research, and in the few short years I have been part of its development it has seen adoption in nanopore labs across the globe, both academic and industrial. Due to the simple nature of CBD, this technique also enables novel applications of nanopores in device architectures that were inaccessible to the expensive and inflexible methods used previously. The second part of this thesis takes advantage of the unique opportunities presented by CBD to develop a device architecture comprising two nanopores in series. This nanodevice tackles one of the main problems standing between nanopores and the promise of cheap genomic analysis: control of the motion and conformation of the polymer both prior to and during translocation through the pore. Finally, because the field of nanopore research is still relatively young, very few tools are available which provide high-quality analysis of nanopore data. The last part of this thesis is dedicated to a thorough discussion of the complexities involved in analysing nanopore signals, as well as the development of several tools which directly address this knowledge gap.

abstract: While techniques for reading DNA in some capacity has been possible for decades, the ability to accurately edit genomes at scale has remained elusive. Novel techniques have been introduced recently to aid in the writing of DNA sequences. While writing DNA is more accessible, it still remains expensive, justifying the increased interest in in silico predictions of cell behavior. In order to accurately predict the behavior of cells it is necessary to extensively model the cell environment, including gene-to-gene interactions as completely as possible. Significant algorithmic advances have been made for identifying these interactions, but despite these improvements current techniques fail to infer some edges, and fail to capture some complexities in the network. Much of this limitation is due to heavily underdetermined problems, whereby tens of thousands of variables are to be inferred using datasets with the power to resolve only a small fraction of the variables. Additionally, failure to correctly resolve gene isoforms using short reads contributes significantly to noise in gene quantification measures. This dissertation introduces novel mathematical models, machine learning techniques, and biological techniques to solve the problems described above. Mathematical models are proposed for simulation of gene network motifs, and raw read simulation. Machine learning techniques are shown for DNA sequence matching, and DNA sequence correction. Results provide novel insights into the low level functionality of gene networks. Also shown is the ability to use normalization techniques to aggregate data for gene network inference leading to larger data sets while minimizing increases in inter-experimental noise. Results also demonstrate that high error rates experienced by third generation sequencing are significantly different than previous error profiles, and that these errors can be modeled, simulated, and rectified. Finally, techniques are provided for amending this DNA error that preserve the benefits of third generation sequencing. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2017

One of the most crucial steps towards nanopore-based nucleic acid analysis is extending the dwell time of DNA molecules within the sensing region of the nanopore. I address this issue by interfacing solid-state nanopores with gel media, which sterically hinders translocating DNA molecules, increasing dwell times. Specifically, my experimental results focus on two reptation regimes: when the DNA molecule is flexible on the length scale of the gel pore, and when the DNA molecule is inflexible on the length scale of the gel pore. The first regime is achieved through the use of agarose gel and 5 kbp dsDNA fragments, and produces a wide distribution of translocation times, spanning roughly three orders of magnitude. The second regime is achieved through the use of polyacrylamide gel and 100 bp dsDNA fragments, and displays a shift in translocation times by an order of magnitude while maintaining a tight distribution.

The miniaturization afforded by the integration of microfluidic technologies within lab-on-a-chip devices has greatly enhanced analytical capabilities in several key applications. Microfluidics has been utilized in a wide range of areas including sample preparation and analysis, DNA microarrays, cell detection, as well as environmental monitoring. The use of microfluidics in these applications offer many unique advantages: reduction in the required sample size, reduction in analysis time, lowered cost through batch fabrication, potentially higher throughput and the vision of having such devices used in portable systems. Nanopore sensors are a relatively new technology capable of detection and analysis with single-molecule sensitivity, and show promise in many applications related to the diagnosis and treatment of many diseases. Recently, some research groups demonstrated the integration of nanopores within microfluidic devices to increase analytical throughput. This thesis describes a methodology for integrating nanopore sensors within microfluidic devices with the aim of enhancing the analytical capabilities required to analyze biomolecular samples. In this work, the first generation of an integrated nanopore-microfluidic device contained multiple independently addressable microfluidic channels to fabricate an array of nanopore sensors using controlled breakdown (CBD). Next, for the second generation, we added pneumatic microvalves to manipulate electrical and fluidic access through connected microfluidic channels. As a proof-of-concept, single molecules (single- and double-stranded DNA, proteins) were successfully detected in the devices. It is also demonstrated that inclusion of the microfluidic via (microvia) limited the exposed area of the embedded silicon nitride membrane to the solution. This helped in localizing nanopore formation by confining the electric field to specific regions of the insulating membrane while significantly reducing high frequency noise in the ionic current signal through the reduction of chip capacitance. The devices highlighted in this thesis were designed and fabricated using soft lithography techniques which are available in most biotechnology laboratories. The core of this thesis is based on two scientific articles (Chapters 3 and 4), which are published in peer-reviewed scientific journals. These chapters are preceded by an introductory chapter and another chapter detailing the experimental setup and the methods used during the course of this study.