No description available.

L’objectif du présent projet est de développer une méthodologie pour la modélisation fondamentale de procédés de polymérisation en émulsion sans tensioactif stabilisés par des particules inorganiques, dénommées "polymérisation en émulsion Pickering". La modélisation des systèmes de polymérisation en émulsion nécessite la modélisation de la distribution de taille des particules (PSD), qui est une propriété importante d'utilisation finale du latex. Cette PSD comprend des sous-modèles dédiés à la nucléation des particules, le transfert de masse entre les différentes phases (monomère, radicaux, stabilisant) et la coagulation des particules. Ces modèles devraient de préférence être validés expérimentalement de manière individuelle. La première partie principale du travail est consacrée à l'étude expérimentale. Cette partie peut être divisée en trois parties. La première partie décrit l'adsorption de particules inorganiques sur le polymère sans réaction. Une adsorption multicouche a été observée et l’isotherme B.E.T. a été capable de décrire cette adsorption. L'adsorption se révèle être plus importante pour une force ionique plus élevée. La dynamique d'adsorption semple être rapide et par conséquent le partage peut être considéré à l'équilibre pendant la polymérisation. La deuxième partie concerne l'étude de différents paramètres de réaction sur le nombre de particules et la vitesse de réaction dans des polymérisations ab initio. L'effet du mélange, de la concentration initiale des monomères et de la concentration de l'initiateur a été étudié. L'optimisation de ces conditions a été utile pour la partie de modélisation. La dernière partie décrit les différences entre plusieurs Laponite® à travers la polymérisation en émulsion ab initio du styrène.La deuxième partie principale du manuscrit a porté sur la modélisation de la polymérisation en émulsion Pickering. Le modèle de bilan de population et le nombre moyen de radicaux par particule ont été calculés en fonction de l'effet des particules organiques. La croissance des particules de polymère a été optimisée en ajustant les modèles d'entrée et de désorption des radicaux décrits dans la littérature aux données expérimentales. Aucune modification n'a été nécessaire, ce qui nous a permis de conclure que l'argile n'avait aucune influence sur l'échange radical. Cependant, la stabilisation joue un rôle important dans la production de particules de polymère. Le modèle de nucléation coagulante a été capable de décrire le taux de nucléation et de prédire le nombre total de particules / The aim of the present project is to develop a methodology for fundamental modeling of surfactant-free emulsion polymerization processes stabilized by inorganic particles, referred to as “Pickering emulsion polymerization”. Modeling emulsion polymerization systems requires modeling the particle size distribution (PSD), which is an important end-use property of the latex. This PSD includes submodels dedicated to particle nucleation, mass transfer between the different phases (monomer, radicals, stabilizer), and particle coagulation. These models should preferably be individually identified and validated experimentally. The first main part of the work is dedicated to the experimental study. This part can be divided in three parts. The first part describes the adsorption of inorganic particles on polymer without reaction. Multilayer adsorption was observed and B.E.T. isotherm was able to describe this adsorption. The adsorption was found to be enhanced at higher ionic strength. The adsorption dynamics were found fast and therefore clay partitioning can be considered at equilibrium during polymerization. The second part concerned the investigation of different reaction parameters on the particles number and reaction rate in ab initio polymerizations. The effect of mixing, initial monomer concentration and initiator concentration were considered. Optimization of these conditions was useful for the modeling part. The last part described the differences between several LaponiteR_ grades through the ab initio emulsion polymerization of styrene. The second main part of the manuscript focused on the modeling of the Pickering emulsion polymerization. The population balance model and average number of radicals balance were adapted regarding the effect of inxi organic particles. The growth of the polymer particles was optimized by fitting the models of radicals’ entry and desorption described available in literature to the experimental data. No modification was needed, which allowed us to conclude that the clay had no influence on radical exchange. However, LaponiteR_ stabilization played an important role in polymer particles production. Coagulative nucleation model was able to describe the nucleation rate and predict the total number of particles

Les piles à combustible, alimentée par de l’hydrogène, représentent une solution prometteuse pour limiter la pollution. L’une des alternatives économiques envisagées à court et moyen terme est de produire l’hydrogène à partir d’un carburant tel que le méthane ou le bio-éthanol. Cette transformation a pour objectif d’obtenir un mélange de gaz riche en hydrogène avec une très faible teneur en CO, ce dernier étant un poison pour les piles de type PEM. La réaction de Water-Gas Shift (WGS) est une étape clé du procédé ; elle convertit CO en CO2 par réaction avec l’eau et fournit une quantité d’hydrogène supplémentaire. Des catalyseurs métalliques (Pt, Pd, Ru, Rh, Au, Cu) supportés sur des oxydes (CeO2, TiO2, ZrO2, Fe2O3, CeO2/Al2O3) ont été comparés dans des conditions de WGS identiques en présence de CO2 et H2. Une étude cinétique a été réalisée sur les catalyseurs Pt/CeO2, Au/CeO2, Pt/TiO2 et Au/TiO2. Les énergies d’activation apparentes et les ordres de réaction ont été déterminés à partir d’un modèle de type loi de puissance. Un mécanisme réactionnel avec deux sites a été proposé pour décrire les différentes activités des 4 catalyseurs. Des expériences de désorption programmée en température ont été réalisées pour déterminer les paramètres cinétiques sur le support / The Fuel Cells are promising solution to reduce the air pollution. One of the cost-efficient alternatives is to produce hydrogen from another fuel such as methane or bio-ethanol. A hydrogen fuel processor consists in generating a hydrogen-rich mixture and reducing the carbon monoxide content, as PEM fuel cells are very low CO tolerance. One of these units is the water-gas shift reactor, which converts CO into CO2 by the reaction with water and provides additional hydrogen. Catalysts based on a metal (Pt, Pd, Ru, Rh, Au, Cu) supported on an oxide (CeO2, TiO2, ZrO2, Fe2O3, CeO2/Al2O3) were compared for the WGS reaction in the same conditions and in the presence of CO2 and H2. A kinetic study was conducted on catalysts Pt/CeO2, Au/CeO2, Pt/TiO2 and Au/TiO2. A power law rate model was used to determine apparent activation energies and reaction orders. A dual-site reaction mechanism was proposed to explain the different activities between the four catalysts. The sorption parameters of H2O and CO2 on the supports was quantitatively determined from temperature-programmed desorption experiments

Ce travail visait à évaluer le transport d’hydrocarbures aromatiques polycycliques (HAP) et de métaux dans des sols multi pollués, en s’intéressant plus particulièrement au couplage transport-réaction dans la zone non saturée du sol, selon deux axes de travail. En premier lieu, puisque de nombreux sites industriels sont multi pollués, nous avons choisi d’étudier l’influence des métaux sur la sorption des HAP et vice versa en système modèle. Nous avons montré que lors d’une pollution simultanée par du fluoranthène et du zinc, ce dernier est beaucoup moins retenu que lorsqu’il est présent seul et peut donc migrer plus facilement vers la nappe phréatique. Le plomb, quant à lui, est plus fortement retenu en présence de fluoranthène. En revanche, l’effet des métaux sur la rétention du fluoranthène est plus limité. En second lieu, nous avons évalué le transport des HAP et des métaux dans une terre d’une ancienne cokerie lorraine à plusieurs échelles : en colonne de laboratoire en conditions saturées et non saturées ainsi qu’en lysimètre de terrain, afin d’évaluer le pouvoir prédictif des sytèmes de laboratoire. Nous avons montré que les expériences en colonne de laboratoire surestiment la sortie des HAP telle qu’elle a lieu sur le terrain. En revanche, la sortie des métaux est bien prédite par de telles expériences. Mais le lysimètre ne peut pas être utilisé lors d’études de risque puisqu’il demande des durées d’expériences de plusieurs années afin d’avoir une vue globale des phénomènes qui s’y produisent. C’est pourquoi nous avons conçu un système original au laboratoire, capable d’évaluer leur transport conformément à ce qui se passe sur le terrain / PAHs and heavy metals are major soil pollutants and most of former industrial soils are co-polluted by both types of compounds. The aim of this work was to assess their transport in polluted soils under saturated and unsaturated flow conditions. This study was carried out by two approaches. First, heavy metal influence on PAH sorption and vice versa was studied with a “model” system. We showed that when the soil was simultaneously contaminated with fluoranthene and zinc, zinc retention was lower than when it was alone in the soil. As a consequence zinc would be expected to migrate more easily to the water table in co-polluted soils. Lead was more strongly retained in the presence of fluoranthene. But heavy metal influence on fluoranthene sorption was of minor importance. Then, PAH and heavy metal transport in a former coking plant soil was evaluated at the laboratory scale, under saturated and unsaturated flow conditions, by means of column experiments, and at the lysimeter scale under natural atmospheric conditions. The goal of these experiments was to compare PAH and heavy metal migration in a field-like situation with results obtained in the laboratory and to determine the appropriate experimental device for risk assessment studies. We showed that column experiments may overestimate PAH migration at the field scale. However heavy metal leaching was in accordance with what was expected from lysimeter experiments. But such experiments are rather time consuming and quite expensive, and as a consequence not adapted for risk assessment studies. Therefore we designed an original laboratory set up able to represent PAH and heavy metal migration at the field scale

Estrogens, alkylphenols and bisphenol-A, enter the environment through waste water systems and waste disposal of manufactured products e.g. detergents, paints, polycarbonates and flameretardants. These analytes disrupt the endocrine function of living organisms affecting their reproductive health and those of future generations. Gas phase low molecular- mass aldehydes and amines are typically eye, nose, and throat irritants. Formaldehyde is classified as a probable human carcinogen. Given their negative impact on human health it is urgent to monitor pollutants at extremely low levels in both air and water. The aqueous pollutants are often concentrated using solid phase extraction cartridges or liquid-liquid extraction followed by derivatization. Methods that can most effectively and selectively pre-concentrate aldehydes and amines involve in situ derivatization. Unfortunately, the derivatizing reagents as well as their associated solvents or adsorbents, are responsible for problems encountered with these methods. Polydimethylsiloxane (PDMS) has emerged as the ideal concentration and reaction medium for trace analysis. However the expensive commercial devices such as SPME and SBSE both require the samples to be returned to the laboratory for concentration. Due to the open tubular nature of the PDMS multichannel trap (MCT), developed in our laboratory, it is ideally suited for on-site and online sampling. The MCTs have a high analyte capacity owing to the large volume of PDMS available for concentration. The derivatization reaction can be performed in situ providing a “onepot concentration and reaction device”. This allows for reduced risk of contamination of / or losses of the sample and a sampling method that can cater for both air and water samples. To demonstrate the versatility of the PDMS MCT, two approaches for concentration in PDMS were investigated in this study, namely, 1) the on-line concentration and in situ derivatization of volatile polar analytes from air followed by REMPI-TOFMS detection, and 2) the concentration of phenolic lipophilic analytes from water requiring derivatization prior to analysis by GC/MS. 1) Analyte and derivatizing reagent were simultaneously introduced into the PDMS trap using a ypress- fit connector. The reaction occurs in situ followed by thermal desorption using a thermal modulator array alone or in conjunction with a thermal desorption unit. The aldehydes and amine derivatives were successfully detected by the REMPI-TOFMS. Reaction efficiencies were determined at room temperature without catalysts. Formaldehyde yielded a low reaction/concentration efficiency of 41 % with phenylhydrazine in PDMS, while acetaldehyde, acrolein and crotonal displayed much improved values of 92, 61 and 74 % respectively. Both propylamine and butylamine yielded 28 % reaction/concentration efficiency with benzaldehyde in the PDMS matrix. Detection limits obtained with this technique were significantly lower than the permissible exposure limits set by the Occupational Safety and Health Administration. It should be noted that the detection limits were not determined by actual measurement but by extrapolation from a larger signal. 2) Aqueous analytes were concentrated in the PDMS MCT using a gravity flow rate of ~50 ìl/min. The trap was dried and 5 ìl derivatizing reagent added. At room temperature and without the presence of a catalyst, the reaction of alkylphenols with trifluoroacetic acid anhydride in the PDMS matrix was 100% complete after 5 minutes. Bisphenol-A reacted less than 50 % to completion during this period, but the amount of derivative formed remained constant. This study revealed that extraction efficiencies of the alkylphenols and bisphenol-A off the PDMS trap have poor batch-tobatch repeatability indicating that the PDMS matrix was not homogenous. For two different PDMS batches: tert-octylphenol displayed an extraction efficiency of 70 and 79%, nonylphenol displayed 84 and 43% while Bisphenol-A displayed 10 and 26% respectively. The thermally desorbed derivatives were analysed by GC/MS. Despite background contamination in the desorption unit, detection limits were at the ppt level. Detection limits were not determined by actual measurement but by extrapolation from a larger signal. / Thesis (PhD)--University of Pretoria, 2009. / Chemistry / unrestricted

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) has revolutionized the field of bacterial diagnostics and is also used for routine analysis in smaller clinical laboratories. For identification of moulds, pre-analytical steps are more complicated and time consuming than for bacteria, and the choice of reference library has a big impact on the utility of MALDI-TOF MS. The aim of this study was to investigate if MALDI-TOF MS is applicable for identification of moulds belonging to the genus Aspergillus at the hospital laboratory in Gävle. Therefore, strains belonging to the genus Aspergillus and Penicillium, were analysed with MALDI-TOF MS after 2, 4 and 7 days of incubation. Two different extraction protocols were used and compared. Mass spectra were compared to reference spectra in two different databases: MSI-2 and RUO Compass library/BDAL (Bruker). Of the strains included, 97 % were correctly identified to species complex level with MSI-2. Only 25 % were identified to species level with RUO Compass library/BDAL (Bruker). However, totally 56 % were correctly identified to species complex level if a lower score value limit than recommended for identification, were applied. Significantly raised score values were observed with one of the protein extraction protocols used. Although, in most cases, the strains were considered identified to species complex level with either method. This pilot study conveys the feasibility of MALDI-TOF MS for identification of Aspergillus species in a clinical laboratory. While there are still issues to address, applying MALDI-TOF MS has the potential to allow for quicker and more precise identification, also in this specific clinical setting.

Volatile and semi-volatile organic compounds are ubiquitous, and some of them are hazardous. The ability to rapidly detect and identify trace levels of them in air has become increasingly important. The conventional device used today for sampling and concentrating them in air is thermal desorption tubes filled with specific sorbents, which can only collect air samples at flow rates of 100-200 mL/min. In order to detect low concentration (ppt level) VOC compounds, long sampling time (>2 h) and sensitive detection are required. At the same time, portable instrumentation for on-site analysis has been developing rapidly. The somewhat lower performance of portable instruments compared to benchtop systems requires the sampling of even greater sample volume in order to reach the same detection limits. In this study, two high flow rate air sampling devices, i.e., a multi-capillary trap and a concentric packed trap, were developed to sample a large volume of air in a short time period. The multi-capillary trap was constructed by bundling analytical capillary gas chromatography columns together in parallel. As low as single digit ppt detection limits were reached in less than 25 min with this trap, and as high as 8.0 L/min flow rate was sampled. The simple and compact multi-capillary trap could be easily used with a conventional thermal desorption system to perform high flow rate sampling. A concentric packed high flow rate trap was also developed by packing sorbent layers concentrically around an empty tube. The concentric packed trap achieved a high flow rate (>10 L/min) because it had a high surface area and short sorbent bed. Also, its large sorbent amount (>1 g) provided large breakthrough volume (>100 L) required to achieve low detection limits. An equilibrium distribution sampling system was developed by absorbing selected analytes in granular PDMS to provide calibration for on-site instrumentation. Furthermore, a needle trap device was coupled in tandem to both high flow rate air samplers to perform second-stage concentration of VOCs down to the ppt level. Concentration factors of 104 to 105 were achieved within 30 min using both systems, i.e., over 10 to 100 times more sample was collected compared to conventional TD systems.

The effect of soot on the catalytic properties of a diesel particulate filter coated with a catalyst for the selective catalytic reduction of NOx with ammonia (SDPF) was studied by means of model-gas experiments. After loading of the SDPF with model soot from 0 to 10 g l-1, the NH3 storage as well as the catalytic DeNOx behavior of the standard SCR reaction was investigated. The model soot present in the filter was shown to have an NH3 storage capacity. The soot deposit inside the SDPF filter wall lead to a decreased NO conversion in SCR experiments of up to 20 %. The NH3 breakthrough was found to be shifted towards earlier time-on-stream during NH3 adsorption on soot loaded SDPF samples. Both effects could be attributed to a diffusive mass transport limitation of the gas species through the soot to reach at the chemically active sites inside SDPF filter wall. The self-diffusion coefficient of NH3 probe molecules within a soot layer could be measured using Pulsed Field Gradient-NMR technique. The unit collector model is capable of describing the backpressure upon soot loading with a depth filtered (inside filter wall) soot amount of 1 g l-1 and 0.36 g l-1, respectively, for both SDPF types under investigation. Based on Scanning Electron Microscopy (SEM) investigation a 1-D microscopic soot filter wall-model was set up. The model implies soot as diffusion barrier for mass transport. It was calibrated based on experimental observations and allows to conclude on the distribution of the soot within the filter wall. Thus, a high soot-coverage of the porous filter wall close to the inlet channel, a slightly covered middle part and a soot free zone close to the outlet explains the observed reduction in NO conversion as well as the NH3 breakthrough at earlier time-on-stream during NH3 adsorption experiments for SDPF samples loaded with soot. A modelled homogeneous soot distribution (0.6 µm soot layer on top of washcoat) within the whole SDPF was shown to result in NO conversion drop up to 45 %.

Gegenstand vorliegender Arbeit ist die Untersuchung unterschiedlicher Modellansätze zur Beschreibung der Extraktion organischer Schadstoffe aus Böden unter Verwendung von überkritischem Wasser. Dazu wurden in der Literatur vorhandene Stofftransportmodelle herangezogen und eigene Modellansätze entwickelt. Das Modell berücksichtigt die Geschwindigkeits- und Temperaturverteilung im Reaktor und berechnet daraus, in Abhängigkeit der desorptiven, diffusiven und konvektiven Stofftransportvorgänge eine Schadstoffverteilung. Zur Lösung der Modellgleichungen wurde FEMLAB 3.1 verwendet. Zur Parameterbestimmung und Modellüberprüfung wurden Extraktionsversuche an real kontaminierten Böden und an künstlich kontaminierten Modellböden durchgeführt. Einfache und komplexe chemische Gleichgewichte der organischen Schadstoffe wurden auf Basis der Gibbs Energetik mit FACTSAGE 5.2 ermittelt und dienten als weitere Inputparameter für die Modellberechnungen.

>Magister Scientiae - MSc / Membrane Capacitive Deionisation (MCDI) is a technology used to desalinate water where a potential is applied to an electrode made of carbonaceous materials resulting in ion adsorption. Processes and materials for the production of electrodes to be applied in Membrane Capacitive Deionisation processes were investigated. The optimal electrode composition and synthesis approached was determined through analysis of the salt removal capacity and the rate at which the electrodes absorb and desorb ions. To determine the conductivity of these electrodes, the four point probe method was used. Contact angle measurements were performed to determine the hydrophilic nature of the electrodes. N2 adsorption was done in order to determine the surface area of carbonaceous materials as well as electrodes fabricated in this study. Scanning electron microscopy was utilised to investigate the morphology. Electrodes were produced with a range of research variables; (i) three different methods; slurry infiltration by calendaring, infiltration ink dropwise and spray-coating, (ii) electrodes with two different active material/binder ratios and a constant conductive additive ratio were produced in order to find the optimum, (iii) two different commercially available activated carbon materials were used in this study (YP50F and YP80F), (iv) two different commercially available electrode substrates were utilised (JNT45 and SGDL), (v) different slurry mixing times were investigated showing the importance of mixing, and (vi) samples were treated at three different temperatures to establish the optimal drying conditions. Through optimization of the various parameters, the maximum adsorption capacity of the electrode was incrementally increased by 36 %, from 16 mg·g-1 at the start of the thesis to 25 mg·g-1 at the end of the study.