Les procédés multi-colonnes de chromatographie ont connu depuis quelques années un développement tel qu'ils sont devenus des standards industriels à toutes échelles, depuis celle des produits pharmaceutiques à haute valeur ajoutée jusqu'à celle des grands intermédiaires chimiques. La spécificité du présent travail consiste à étudier, pour ces procédés, l'influence d'un gradient d'élution. Il s'agit de faire varier au cours du temps la force éluante de la phase mobile. L'objectif est d'augmenter la productivité et le taux de récupération d'un produit à haute valeur ajoutée, tout en répondant à des contraintes de pureté. L'utilisation d'un gradient de solvant, courante en chromatographie analytique, fait l'objet d'un intérêt plus récent en chromatographie préparative. Les applications visées concernent des séparations de mélanges complexes où l'espèce cible a une affinité intermédiaire pour le support solide par rapport à celle des autres espèces, ce qui est souvent le cas lors de la purification de biomolécules issues de matières premières naturelles ou issues des biotechnologies. Dans ce cas, la séparation conduit à trois fractions, des impuretés faiblement retenues, la fraction intermédiaire et des impuretés fortement retenues. Pour notre étude, un mélange modèle, peu coûteux et non toxique, de cinq acides aminés a été choisi. Ces acides aminés ont été choisis en tenant compte de leur caractère apolaire et hydrophobe. Les séparations ont été réalisées par chromatographie en phase inverse. Dans un premier temps, une étude expérimentale, réalisée à l'aide d'une chaîne HPLC, a permis de déterminer les paramètres des isothermes d'adsorption de chaque acide aminé pour différentes teneurs en solvant organique de l'éluant. Une loi empirique a permis de relier le facteur de rétention k à la composition de la phase mobile (K = f (xméthanol)). Un travail de modélisation/simulation, reposant sur l'approche d'une cascade de mélangeurs, a ensuite permis de simuler les séparations obtenues dans le cas d'une seule colonne, puis dans le cas d'un système multi-colonnes. L'utilisation des lois reliant les facteurs de rétention k à la concentration en modifieur a alors permis de réaliser des simulations pour différents gradients de solvants. Dans le cas d'une seule colonne, le gradient a été optimisé en minimisant la durée de la séparation et en respectant une contrainte sur la résolution des pics des 2 espèces les plus difficiles à séparer. Une bonne adéquation a été observée entre les simulations et les résultats expérimentaux obtenus avec un gradient sur une seule colonne. Des expérimentations numériques ont alors été réalisées dans le cas du système multi-colonnes. Les paramètres opératoires optimaux ont été déterminés dans le cas du mélange étudié. Ces réglages seront ainsi utilisés lors de la validation expérimentale qui sera réalisée sur l'unité pilote. Cette unité comporte trois colonnes. Il s'agit d'un procédé séquentiel cyclique. Pour le mode opératoire retenu, chaque cycle comporte 8 étapes. A chaque étape les alimentations et soutirages des différentes colonnes sont modifiées. Pour le soutirage qui correspond à la fraction de l'espèce cible, les critères étudiés seront la pureté et le taux de récupération / Multi-column chromatographic processes have known, for a few years, a development on all scales, from high added value pharmaceutical products to major chemical intermediates. The specificity of the present work is to study the influence of a gradient elution for these processes. It consists in varying the eluent strength of the mobile phase over the time. The aim is to increase the productivity and the recovery ratio of a high added value product, while satisfying the constraints of purity. Solvent gradient is currently used in analytical chromatography and presents a recent interest in preparative chromatography. The applications concern separations of complex mixtures where the target species has an intermediate affinity for the solid phase compared to other species, which is often the case during the purification of biomolecules extracted from natural raw materials or resulting from biotechnologies. In this case, separation leads to three fractions, impurities weakly retained, an intermediate fraction and impurities strongly retained. For our study, a model mixture, inexpensive and nontoxic, of five amino acids was selected considering their nonpolar and hydrophobic character. The separations were carried out by reversed phase chromatography. An experimental study using a HPLC system was first carried out with single-element solution of each amino acid in isocratic mode. This enabled to determine adsorption isotherm parameters. An empirical law giving the retention factor as a function of eluent composition was determined (K = f (xmethanol)). A work of modeling / simulation, assuming linear isotherm and based on the mixed cells approach, permitted to simulate the separations obtained in the case of a one-column process, then in the case of a multi-column system. The use of retention factors laws allowed to carry out simulations for different solvent gradients. In the case of a single column, a simple methodology was developed to calculate the optimal solvent gradient. The gradient was optimized by minimizing the separation time and by respecting a constraint on the peaks resolution of the two species which are the most difficult to separate. A really good adequacy was observed between simulations and the experimental results. Numerical experimentations, executed in the case of the multi-columns process, made it possible, yet, to find the optimal operating parameters in the case of the studied mixture. These settings will be applied in the experimental validation which will be realized on the pilot unit. This unit has three columns. It is a cyclic sequential process. For the selected operating mode, each cycle contains eight steps. At each step, inlets ant outlets streams of different columns are switched. The criteria for the target species fraction are purity and recovery

Orientador: Carol Hollingworth Collins / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Quimica / Made available in DSpace on 2018-08-07T20:37:42Z (GMT). No. of bitstreams: 1 Faria_AnizioMarciode_D.pdf: 2871302 bytes, checksum: 58d9a466b2a4724df83a92414de3d9a9 (MD5) Previous issue date: 2006 / Doutorado / Quimica Analitica / Doutor em Quimica

Introduction Oligonucleotides are short nucleic acid chains, usually 19-27mer long. They bind to their corresponding chain, making a specific inhibition possible. In pharmaceuticals, this can be used to inhibit the expression of a gene or protein of interest. Oligonucleotides are usually analyzed based on separation using both hydrophobic and ion-exchange properties. In this project, the possibility to use a mixed-mode column to separate these oligonucleotides and their impurities were explored. Method Liquid chromatography is used as the separation method and the method of detection is both mass spectrometry and UV. Three different columns are evaluated; C18, DNAPac RP, and mixed-mode RP/WAX. Results and discussion Different compositions of mobile phases and gradients are evaluated based on a literature study. Triethylamine, triethylammonium acetate, ammonium formate, hexafluoroisopropanol is used along with both methanol and acetonitrile. Phosphate buffer is evaluated on LC-UV. The results from the C18 column displays a good separation of the oligonucleotides, whilst the DNAPac RP is not as sufficient using the same mobile phases. The mixed-mode column provides good separation and selectivity using phosphate buffer and UV detection. Conclusion Mixed-mode column has the potential to be used for separation of oligonucleotides and one future focus would be to make the mobile phase compatible with mass spectrometry. Phosphate buffer and UV detection seems to be the go-to mobile phase using mixed-mode column even though MS is a more powerful tool for the characterization and identification of oligonucleotides. This provides a hint about the challenge in making the mobile phase MS compatible.

<p>Liquid chromatography coupling with mass spectrometry (LC/MS) plays an important role in pharmaceutical characterization because of its ability to separate, identify, and quantify individual compounds from the mixture. Polymer brush layer bonded to the silica surface is designed as a novel stationary phase to improve the LC resolution and MS compatibility. The polymer thickness can be controlled to shield the analyte from interacting with the active silanol on the surface and reduce peak tailing. The functional group of the polymer can be changed to tune the selectivity in different separation modes. </p><p> </p><p>Two projects on LC/MS method development for biomolecule characterization using polymer-shell column are discussed in this work. In the first project, a polymer-shell column is used for disulfide bonds and free thiol subspecies identification, which is a major type of structural heterogeneities in IgG1. Compared with commercial columns, the polymer-shell column is able to resolve the free thiol variants without the presence of trifluoroacetic acid and greatly improve the MS signal. In the second project, a polymer-shell column is used for characterizing the drug-loading profile for antibody-drug-conjugates (ADC) via online LC/MS. The separation employs a mobile phase of 50 mM ammonium acetate to keep the ADC intact, and a gradient of water/isopropanol for ADC elution. MS data show that all ADC species remained intact and native on the column. Positional isomers can be separated and identified with the new method as well. Furthermore, to understand the surface chemistry and protein separation behavior quantitatively, a chromatographic simulation study is performed. The result shows that protein separation in RPLC can be described by a bi-Langmuir adsorption isotherm with mixed-mode retention of strong and weak sites. Smaller fractions and lower equilibrium constant of the strong site, which is the active silanol, give less tailing for protein separation.</p>

<p>Therapeutic proteins are often lyophilized with excipients such as sucrose or trehalose to protect them during manufacturing and achieve a longer shelf-life. Formulation design for therapeutic proteins has been a trial-and-error process, and the mechanisms responsible for the stabilizing effects of excipients are not fully understood. Two proposed theories have been widely accepted: the water replacement theory and the vitrification theory.^1,2The water replacement theory suggests that excipients stabilize protein molecules in the solid state by forming hydrogen bonds that “replace” the hydrogen bonds to water that stabilize the protein in solution, while the vitrification theory asserts that proteins are stabilized by a glassy solid matrix of low mobility and does not require direct interactions between excipient and protein. A better understanding of the interactions between proteins and other components of the lyophilized matrix can facilitate rational formulation design and shorten the time in development. However, most of the analytical methods available can only provide information on the bulk properties of the lyophilized matrix such as moisture content and glass transition temperature (<i>T</i>_g); it has been difficult to measure the interactions between protein and excipient directly, if they exist. In order to characterize the interactions between protein and excipients in a lyophilized matrix with high resolution, a photolytic labeling method was developed in this dissertation, building on previous work in our research group. Photolytic labeling has long been used to identify protein-protein interactions <i>in vivo</i>.^3,4Common types of photo-reaction reagents and their applications are summarized in Chapter 1. The research described in this dissertation utilizes the diazirine functional group, which is activated after UV exposure and undergoes a free radical reaction to form covalent bonds with nearby molecules. The reaction can be used to identify the interactions between excipients and protein or peptide in a solid formulation. Previous studies in our lab have shown that photo-reaction can be applied to lyophilized solids to study protein-matrix properties and interactions in the solid.^5,6This dissertation seeks to further identify photo-reaction products and analyze them in a more quantitative way. </p><p> </p><p>Chapter 2 describes a quantitative analysis of photo-reaction products in solution and lyophilized solids using a model peptide, KLQ (Ac-QELHKLQ-NHCH₃). The purpose of the work in this chapter is to establish a quantitative analytical method for photo-reaction products, enabling studies of peptide-excipient interactions in lyophilized solids. KLQ was derivatized with a bifunctional probe NHS-diazirine (succinimidyl 4,4’-azipentanoate; SDA) at Lys5 to be photo-reactive. The SDA derivatized KLQ (KLQ-SDA) was used to study the photo-reaction products and examine excipient interactions. Identification and quantitation of photo-reaction products of KLQ-SDA was achieved with liquid chromatography mass spectrometry (LC-MS) and reversed phase HPLC (rp-HPLC). Important reaction products such as peptide-excipient adducts and peptide water adducts varied in different formulations. Unexpected reaction products such as unproductive “dead-end” products and peptide-phosphate adducts from buffer salt were also detected and quantified. Together, the photo-reaction products reflected the local environment near Lys5 of the peptide in the solid state. This study has provided a better understanding of photo-reaction with diazirine in the lyophilized solids together with a quantitative description of the local environment near Lys5. </p><p> </p><p>In Chapter 3, the photo-reaction products in lyophilized solids exposed to increasing moisture were analyzed, and the effect of increasing moisture on the local environment near the peptide was examined. Using the analytical method developed in Chapter 2, these studies explored whether peptide-water interactions, as measured by the formation of water adducts formed by photolytic labeling, are linearly correlated with an increase in solid bulk moisture content. Formulations containing the KLQ-SDA peptide were exposed to various relative humidity conditions and photolytic labeling was induced. Solids containing disaccharide excipients behaved differently from those containing amino acids when exposed to the same relative humidity condition, showing different levels of peptide-excipient and peptide-water adducts. With increasing moisture content in the solids, the formation of photo-reaction products did not mimic the pattern of solutions with same composition, indicating differences in the local environment. </p><p> </p><p>An alternative approach to studying lyophilized formulations using photolytic labeling is to incorporate photo-reactive excipients into the solid matrix. In Chapter 4, a new diazirine-labeled photo-excipient, photo-glucosamine (pGlcN), was chemically synthesized and incorporated into formulations of the therapeutic peptide salmon calcitonin (sCT) and compared with the commercially available diazirine-labeled amino acid, photo-leucine (pLeu). The studies in Chapter 4 further compared peptide-excipient interactions at the molecular level with two different photo-excipients, ionizable pLeu and unionizable pGlcN. Changing solution pH prior to lyophilization was expected to change ionic interactions between sCT and pLeu in the solid samples, resulting in different distributions of photo-reactions products; pH-dependent differences were not expected for pGlcN. The results demonstrated that the distribution of photo-reaction products varied with the composition of the formulation and the pH of the solution prior to lyophilization. The photo-reaction products in the pGlcN-containing formulation differed from those pLeu, showing a difference in the interactions of unionizable (pGlcN) and ionizable (pLeu) excipients with sCT in solid samples. </p><p> </p><p>The work in this dissertation has developed photolytic labeling as a tool to study lyophilized peptide formulations, and has provided a more quantitative understanding of the photo-reaction products that are produced from diazirine-labeled peptides or excipients in the solid state. A new photo-reactive excipient has also been presented (pGlcN), which showed different photo-reaction products than a commercially available photo-excipient (pLeu) and is promising for future study. Photolytic labeling for formulation development is still in its early stages, and additional research regarding reaction mechanism and complementary stability studies is needed. Nevertheless, the results presented in this dissertation support continued development of photolytic labeling as a practical method for formulation design and development. </p><p> </p>

Magnetically confined fusion (MCF) plasmas are typically subject to several unstable modes. The growth of one mode can limit the plasma energy confinement and might cause a termination of the plasma. Externally applied resonant magnetic perturbations (RMPs) are used to control and to mitigate some of the unstable modes. Examples are, mitigation of edge localized modes and steering of neoclassical tearing mode position for stabilization by electron cyclotron current drive. Consequently, use of RMPs are considered necessary in planned future fusion machines. There are however negative consequences, the RMP interaction with a tearing mode (TM) of the same resonance can cause deceleration of the TM and possibly wall-locking. If a TM is non-rotating relative the machine-wall, it can grow and degrade fusion plasma performance and lead to a plasma disruption. Thus, all fusion confinement machines want to avoid wall-locked modes. Resonant magnetic fields can also be present in the form of machine-error-fields, which can produce the same effects. Clearly, it is of importance to understand the TM-RMP interaction. Typically, the modes with long wavelength are described by magnetohydrodynamic (MHD) theory. Considering the finite plasma resistivity, MHD predicts a mode that tears and reconnects magnetic field lines, called a tearing mode (TM). TMs occur at surfaces where the magnetic field lines close on themselves after a number of (m) toroidal and (n)poloidal turns. These surfaces are resonant in the sense that magnetic field and helical current perturbation has the same helicity, which minimize stabilizing effect of magnetic field line bending. In this thesis, the mechanisms of TM locking and unlocking due to external resonant magnetic perturbations (RMPs) are experimentally studied. The studies are conducted in two MCF machines of the type reversed-field pinch (RFP): EXTRAP T2R and Madison Symmetric Torus (MST). The studied machines exhibit multiple rotating TMs under normal operation. In EXTRAP T2R TM locking and unlocking are studied by application of a single harmonic RMP. Observations show that after the TM is locked, RMP amplitude has to be reduced significantly in order to unlock the TM. In similar studies in MST unlocking is not observed at all after turn-off of the RMP. Hence, in both machines, there is hysteresis in the locking and subsequent unlocking of a tearing mode. Results show qualitative agreement with a theoretical model of the TM evolution when subjected to an RMP. It is shown that the RMP cause a reduction of TM and plasma rotation at the resonant surface. The velocity reduction is opposed by a viscous torque from surrounding plasma. After TM locking, relaxation of the whole plasma rotation is observed, due to the transfer of velocity reduction via viscosity. This results in a reduced viscous resorting torque, which explains the observed hysteresis. The hysteresis is further deepened by the increase in amplitude of a locked mode. / <p>QC 20160111</p>

Upcycling of textiles is a well-known method to remake worn textiles and decrease the environmental impact coming from the textile industry. Many fast fashion companies have tried to implement upcycling in their own textile value chains to become more circular. Although the effort is good in theory, there is more to be done to get better efficiency when it comes to upcycling textile fashion products. The purpose of this report was to study the obstacles in the textile upcycling chain for small-scaled actors that are engaged in design driven upcycling. This report partly investigates a specific textile value chain that is based on redesigning home textiles that comes from textile consumer waste. The report investigates the collaboration between supplier, designer and manufacturer in this specific value chain. This was by outlining their current communication and process steps. The report intends to form a ground for creation of communication tools for actors working similarly. Re:textile at Science Park Borås is a project that is working towards developing new design principles, business models and production systems in the textile industry for a better circular flow system. The cooperation with Anna Lidström, Artistic Director at Re:textile, made it possible to investigate the upcycling industry and identify the obstacles in the pre-production and production process. For collecting information for this report, data research and interviews with Swedish companies have been done. The companies that attended the interviews were from different companies in the textile upcycling industry, that were provided with as much information as possible for the analysis. The companies that attended the interviews were from different companies in the textile upcycling industry. The companies were Rave Review, XV Production, Björkåfrihet and SIPTex. The information from the interviews contributed a ground for the analysis. The semi-structured interviews varied from telephone interviews and visits. The compilations from the interviews have been carefully used to answer the reports research questions. Conclusions are presented as obstacles for the textile upcycling chain at design driven small-scaled actors. One obstacle is to cater bigger quantities of textile consumer waste that keeps the same quality. Another obstacle is the way of ensuring the quality of the product through the value chain to consumers. The third obstacle is that the communication tools are not written according to any principles which contribute to mistakes in the manufacturing process. / Återbruk av textilier är en väl omtalad metod för att ta vara på använda textilier för att minska på miljöpåverkan som kommer från textilindustrin. Många snabbt modeväxlande företag har försökt implementera återbruk i deras egna värdekedjor för att bli mer cirkulära. Även om tanken är god, så är det mer som behöver göras för att få en bättre effektivitet när det kommer till att återbruka textila modeprodukter. Syftet med rapporten var att undersöka hinder i den textila återbrukskedjan hos småskaliga aktörer som ägnar sig åt designdriven återbruk. Rapporten utreder dels en specifik textil värdekedja som grundar sig på att omdesigna hemtextilier som kommer ifrån textilt konsumentavfall. Rapporten utreder samarbetet mellan leverantör, designer och tillverkare i denna specifika värdekedja. Detta genom att redogöra för deras nuvarande kommunikation och processteg. Studien avser att ligga till grund för skapandet av kommunikationsverktyg för aktörer som arbetar liknande. Re:textile på Science Park Borås är en verksamhet som jobbar med att utveckla nya designmetoder, företagsmodeller samt produktionssystem i textilindustrin för ett bättre cirkulärt flödessystem. Samarbetet med Anna Lidström, Konstnärlig Ledare på Re:textile gjorde det möjligt för författarna att undersöka återbruksindustrin och identifiera bristerna i förproduktion och produktionsprocessen. För insamling av information till denna rapport har datainsamling och intervjuer med svenska företag gjorts. Företagen som ställde upp på intervju var ifrån olika verksamheter i den textila återbruksindustrin. Företagen var Rave Review, XV Production, Björkåfrihet och SIPTex. Information från intervjuerna bidrog till underlag för en analys. De semi-konstruerade intervjuerna varierade med både telefonintervju samt platsbesök. Sammanställningarna från intervjuerna har med aktsamhet använts för få svar på rapportens frågeställningar. Slutsatser redogörs som hinder för den textila återbrukskedjan hos designdrivna småskaliga aktörer. Ett hinder är bland annat att tillgodose större kvantiteter av textilt konsumentavfall som håller samma kvalitet. Ett annat hinder är att säkerställa kvalitén av produkten genom värdekedjan till kund. Ett tredje hinder är att kommunikationsverktygen inte är skrivna enligt några principer vilket bidrar till misstag i tillverkningsprocessen.

Le retour haptique disponible aujourd'hui dans les produits grand public est d'un intérêt limité pour les interactions tactiles et moins efficace que l'utilisation d'un clavier physique pour la saisie de texte. Relativement simple, celui-ci ne peut communiquer que peu d'informations : signaler silencieusement un appel, notification de messages ou confirmation de frappe de touches sur clavier virtuel. Bien que des améliorations aient été apportées aux technologies haptiques existantes, comme des actionneurs plus performants et des gammes de vibrations plus larges afin de simuler des boutons ou des textures, elles restent limitées à un retour tactile unique. Ceci empêche tout usage multi-doigts ou multi-utilisateurs en simultané.Ce travail vise à développer un retour tactile statique et dynamique sur grande surface (format A4). Les interactions avec les écrans tactiles nécessitant un retour tactile plus riche et plus performant, deux types de retour complémentaires ont été identifiés afin de les enrichir. Le retournement temporel des ondes de flexions dans les plaques est étudié afin de simuler l'appui sur un bouton (retour statique). La 2ème approche se base sur la stimulation par électrovibration, qui permet de simuler des textures ou de différencier des zones d'interactions (retour dynamique). Afin d’estimer de manière précise la résolution spatiale du procédé tactile par retournement temporel, un modèle analytique basé sur l'équation de Kirchhoff est proposé. Des mesures expérimentales confrontées au modèle ont permis de le valider. Par ailleurs, des règles de conception sont élaborées et appliquées pour le développement d'un nouveau prototype avec une électronique dédiée sur une plaque en verre de faible épaisseur (1.1 mm). Différents types de signaux de commande sont étudiés. La quantification sur un bit (i.e. signaux de forme carré) avec filtrage des fréquences audibles s’avère être l'alternative la plus efficiente en terme d'amplitude de déplacement générée et de réduction des émissions sonores. Des problématiques de dimensionnement, comme le placement des actionneurs, l'homogénéité de la résolution spatiale et l'amplitude de déplacement sont analysées. L'effet de la force d'appui du doigt sur l'amplitude de déplacement est quantifié (6 % de perte d'amplitude dû à une force d'appui de 2 N sur une localisation autre que le point de focalisation, et jusqu'à 37 % pour la même force d'appui sur le point de focalisation).Le seuil de détection d'une focalisation par retournement temporel mesuré sur 10 utilisateurs se situe à environ 10 µm et est peu influencé par la force d'appui de l'utilisateur sur l'écran. En répétant la focalisation des ondes de manière à former un signal modulé en amplitude, il devient possible de générer des retours tactiles enrichis, notamment de simuler le comportement du clic d’un bouton poussoir. Des motifs avec des fréquences de répétition et des enveloppes différentes sont comparés. Il apparaît qu'une fréquence de 200 Hz et une enveloppe en sinus cardinal sont les plus plaisants pour l’utilisateur.Par ailleurs, l'électrovibration produit des stimuli capables de reproduire une sensation de texture, en modifiant le coefficient de friction entre le doigt et la surface à explorer. L’intensité de ces stimuli dépend de l'épaisseur de peau du bout du doigt. Les seuils de détection des mécanorécepteurs sont dépendants de la fréquence du signal appliqué. Une étude utilisateur ayant pour but de déterminer l'influence de la force d'appui sur le seuil de détection d’une stimulation par électrovibration a été conduite. Les seuils minimaux ont été observés pour une fréquence de 240 Hz. La force d'appui a une influence limitée sur les seuils de détection.La combinaison des deux approches de stimulations (retournement temporel et électrovibration) sur une même surface offre un retour tactile riche et multi-point pour une interaction statique (simulation de clics) et dynamique (simulations de textures). / The current haptic feedback in end user products provides limited tactile interactions and is less efficient than physical keyboards for typing. Most people are used to the simple tactile feedback available in smartphones. However, it is very limited, and can only convey little information: silently signaling a phone call, notifying an incoming message or acknowledging touch inputs when typing on a virtual keyboard. Although advances are made to enrich existing technologies in hand-held devices, such as more efficient actuators with broader ranges of vibrations to emulate buttons or textures, they remain limited to a single point feedback. This prevents any simultaneous multi-user scenario.This work aims to develop static and dynamic haptic feedback on large surfaces (A4 format). Interaction with screen based devices is in need of better and richer haptic feedback. Two types of feedback with complimentary performance are identified as necessary to enrich tactile interactions. Time reversal, as a static feedback technology, is studied to simulate a button press. Electrovibration, as a dynamic feedback, is investigated to simulate tactile textures or to differentiate specific areas of interaction.An analytical model based on Kirchhoff's equation for wave propagation to compute the spatial resolution of time reversal of flexural waves applied to plates is presented. Measurements on a physical system are confronted to the model's prediction. Design guidelines are elaborated and used to develop a new time reversal enabled screen with adapted drive electronics, on a 1.1 mm thick glass plate. Driving signal alternatives are investigated. Signals quantified on one bit (i.e. square type signals) with audible frequencies filtered out are found to be the most efficient in terms of amplitude generation and audible noise emission. Integration issues, such as the actuators’ distribution on the plate and their impact on focalisation point's amplitude and spatial resolution homogeneity are investigated. The effect of the fingertip pressure on the amplitude vibration is studied (6% loss of amplitude due to a 2N force applied by a fingertip on a position other than the focalisation location, and up to 37% for the same force at the focus point's location).The detection threshold measured on ten users is found to be about 10 µm and is not influenced by the force applied on the screen. While a single impact (one impulse) demonstrates the feasibility of time reversal for tactile feedback, a repetition of impacts varying in amplitude offers the possibility to generate richer haptic feedback (such as a button click). Patterns with different repetition frequencies and envelopes are compared in a user study. It appears that frequencies of 200 Hz and the smoothness of the cardinal sine envelope are found to be the best in terms of pleasantness.On the other hand, electrovibration stimulations are able to create a texture feedback by modifying the apparent friction coefficient between the fingertip and the surface. The electrostatic force generation depends on the fingertip skin's thickness. The mechanoreceptors detection threshholds are frequency dependent. A user study on the influence of the applied force on the perception threshold of tactile feedback is presented. The minimum perception thresholds are observed for 240 Hz stimulus. The effect of the applied force appears to have limited effect on the perception threshold.The combination of both stimulation approaches (time reversal and electrovibration) on a single surface will offer a rich multi-point tactile feedback, both for static buttons and dynamic textures.

Reversed-phase liquid chromatography (RPLC) is a widely used technique for analytical separations but routinely requires empirical optimization. Gaining a better understanding of the molecular reasons for retention may mean more efficient separations with fewer trial and error runs to obtain optimized separations. Vibrationally resonant sum frequency generation (VR-SFG) is a surface specific technique that has allowed for in situ examination of model RPLC stationary phases under various solvent and pressure conditions. In order to improve on past work with model RPLC stationary phases two challenges had to be overcome. First, improved vibrational mode assignments of the C18 stationary phase were needed for proper understanding of this model system. Second, the synthesis of back-surface reference mirrors used in these VR-SFG experiments allowed us to better correct the relative intensities of the various spectral peaks present in typical spectra. After examination of model RPLC systems under various conditions, we have found that these model substrates have a significant amount of interference from nonresonant signal. This interference of resonant and nonresonant signals on fused silica surfaces has not been previously examined and further studies of the model RPLC stationary phase must properly deal with the non-negligible nonresonant interference that is present. We have seen changes in the VR-SFG spectra of these model systems under a variety of conditions including elevated pressure, however the changes are mostly due to nonresonant interference. These spectral changes, although apparently not solely from structural changes, need to be investigated further to better understand the molecular basis of retention in model RPLC systems.

My work has focused on a variety of different types of diamond-based, core-shell particles. These particles are formed with inert cores and poly(allylamine)/nanodiamond shells. Their intended purpose is to form an LC stationary phase that is stable from pH 1 – 14 and at elevated temperatures. At the beginning of my studies, the particles that had been made in the Linford laboratory were pH stable, but irregular and had poor mechanical stability. Since that time, I have worked to improve the particles by using more spherical zirconia and carbon cores, and I have improved their mechanical stability via chemical crosslinking with epoxides. I have performed van Deemter and van’t Hoff analyses to understand the properties of these columns. Efficiencies greater than 100,000 N/m are routinely achieved with these carbon/nanodiamond-based phases. In addition I contributed to two patents that show innovations in diamond functionalization. My contributions involved reduction of an oxidized diamond surface with LiAlH4 prior to functionalization with isocyanates. I also wrote some application notes for the Flare mixed-mode column, which was recently introduced to the market and contains particles comprised of a carbon core and a polymer/nanodiamond shell. These application notes show the gradient separations of four essential oils (lavender, melaleuca, peppermint and eucalyptus), and the isocratic separations of various triazine herbicides and a mixture of β2-agonists and amphetamines.This dissertation contains the following sections. Chapter 1 is a review of liquid chromatographic history and theory. It also includes a history of the use of diamonds in liquid chromatography. Chapter 2 is a study on a glassy carbon core - polymer/nanodiamond shell particle made in our laboratory. Stability studies at pH 11.3 and 13 were performed and different analytes were retained and/or separated on the column. Chapter 3 is a study performed on the Flare mixed-mode column. Separations of tricyclic antidepressants, β2-andrenergic receptor agonists, and linear chain alkylbenzenes were demonstrated with this phase. Van Deemter and van’t Hoff studies were also performed to probe the efficiency and selectivity of this column with different classes of analytes. Chapter 4 chronicles, via SEM and van Deemter analysis, the improvements that have taken place in our column after many iterations of improved synthetic methods and new materials. These include better particle uniformity, particle stability, and column efficiency. Three different carbon cores were analyzed, each better than the previous one. Appendices 1 – 6 are application notes published by Diamond Analytics of β2-andrenergic receptor agonists and amphetamines, triazine herbicides, and lavender, melaleuca, eucalyptus and peppermint essential oils. Appendices 7 and 8 are patents that contain ideas and research contributed by the author.