Les travaux de thèse portent sur le développement de méthodes pour la caractérisation des nanoparticules (NPs) de dioxyde de titane (TiO2) dans les aliments. L’étude a été séparée distinctement en deux axes. Le développement de la méthode AF4-ICP-MS (I) au sein de la Plateforme Nano-Sécurité du CEA de Grenoble et le développement de la méthode Sp-ICP-MS (II) au sein du Laboratoire de la Sécurité Aliments de l’Anses à Maisons-Alfort.Pour le premier axe, les interactions électrostatiques entre les particules et la membrane de l’AF4 ont été étudiées. La charge de surface (potentiel zêta) a été mesuré pour différentes type de particule et membrane en fonction de plusieurs paramètres du solvant tel que le pH, la force ionique ainsi que la présence de tensio-actifs. Suite aux nombreuses analyses effectuées plusieurs conditions de solvant ont été identifiées comme favorable à la répulsion électrostatique entre les particules et la membrane ce qui permettrai de diminuer les pertes de particules au niveau de la membrane. Plusieurs conditions ont donc été testées directement pour l’analyse de NPs de TiO2 en AF4-ICP-MS. Il a ainsi été montré que les conditions favorisant les répulsions électrostatiques permettaient d’améliorer le taux dé récupération des particules au niveau du système AF4. Cependant les analyses ont aussi montré que les NPs de TiO2 n’étaient pas suffisamment dispersées pour pouvoir obtenir une distribution en taille fiable.Le second volet de la thèse, a eu pour but, dans un premier temps, d’optimiser le traitement de données de la méthode Sp-ICP-MS. Plusieurs améliorations ont été réalisées comme le calcul automatique du seuil bruit de fond-particule, le calcul de l'efficacité de transport où encore l'ajout de paramètres de contrôle qualité. Ce traitement de données a ensuite été comparée à un logiciel dédié aux analyses par Sp-ICP-MS et a permis de mettre en avant que la feuille de calcul interne donne des résultats avec une meilleure justesse, tout en fournissant des paramètres supplémentaires.Dans un second temps une optimisation de la méthode Sp-ICP-MS a été réalisée. Plusieurs paramètres ont donc été évalués comme le choix de l’isotope de titane, le temps d’acquisition, le type de solvant pour disperser les particules, ainsi que le mode de prélèvement de l’échantillon et son débit.Enfin, plusieurs aliments ont été analysés après avoir mis en place un protocole d’extraction des NPs. Les résultats obtenus ont permis de déterminer une distribution en taille des NPs de TiO2 présentes dans les échantillons. / The thesis focuses on the development of methods for the titanium dioxide (TiO2) nanoparticles (NPs) characterization in food products. The study was separated distinctly into two axes. The development of the AF4-ICP-MS (I) method within the CEA Nano-Safety Platform in Grenoble and the development of the Sp-ICP-MS (II) method within the Food Safety Laboratory of the Anses in Maisons-Alfort.For the first axis, the electrostatic interactions between the particles and the AF4 membrane were studied. The surface charge (zeta potential) was measured for particles and membranes depending on several solvent parameters such as pH, ionic strength and the presence of surfactants. Following the several analyzes carried out, several solvent conditions have been identified as promoting the electrostatic repulsion between the particles and the membrane, which will make it possible to reduce the losses of particles at the level of the membrane. Several conditions were therefore directly tested for NPs analysis of TiO2 in AF4-ICP-MS. It has thus been shown that the conditions favoring electrostatic repulsions make it possible to improve the efficiency of the particles at the level of the AF4 system. However, the analyzes also showed that the TiO2 NPs were not sufficiently dispersed to obtain a reliable size distribution.The second part of the thesis aimed at first optimizing the data treatment of the Sp-ICP-MS method. Several improvements have been made such as the automatic calculation of the background-particle threshold, the calculation of the transport efficiency or the addition of quality control parameters. This data treatment was then compared to a software dedicated to analyzes by Sp-ICP-MS and made it possible to highlight that the internal spreadsheet gives results with a better accuracy, while providing additional parameters.In a second time an optimization of the Sp-ICP-MS method was carried out. Several parameters were therefore evaluated such as the choice of the titanium isotope, the acquisition time, the type of solvent to disperse the particles, as well as the sample collection mode and its flow rate.Finally, several foods were analyzed after setting up an extraction protocol for NPs. The results obtained made it possible to determine a size distribution of the NPs of TiO2 present in the samples.

Les faisceaux de photons produits par les accélérateurs d'électrons linéaires médicaux sont plats, grâce à un cône égalisateur. Les technologies ont évolué et la présence d'un cône n'est plus indispensable. On parle alors de faisceaux FFF (flattening filter free). Les faisceaux FFF présentent des débits de dose plus élevés, des profils de dose hétérogènes, des spectres énergétiques différents et une diminution de la dose hors-champ. Cette thèse a eu pour but d'étudier les caractéristiques des faisceaux FFF, ainsi que l'impact de leur utilisation thérapeutique. Plusieurs thématiques ont été. Des expériences d'irradiation in vitro ont tout d'abord permis de s'assurer que les débits de dose FFF n'ont pas d'impact radiobiologique sur la réponse des cellules irradiées. Une large revue de la littérature a permis de corroborer ces résultats. Afin de maitriser les caractéristiques physiques des faisceaux FFF, des mesures ont été faites avec différents détecteurs. Les effets du spectre et du débit de dose sur la calibration en dose ont aussi été étudiés. Les faisceaux FFF ont été modélisés dans deux TPS. Les modèles ont été comparés entre les deux types de faisceaux et entre les deux TPS. La mise en place des traitements stéréotaxiques a aussi été l'occasion d'appréhender la dosimétrie des petits faisceaux. Nous avons étudié des cas VMAT de cancer de la prostate et des cas de stéréotaxies 3D de tumeurs pulmonaires. La comparaison donne un avantage aux faisceaux FFF. La maitrise de la physique et de la biologie des haut débits a permis de débuter les traitements FFF à l'IPC. Des études comparatives nous permettent aujourd'hui d'adapter leur utilisation au cas par cas. / In medical linear electron accelerators, photon beams profiles are homogenised using flattening filters. Technologies have evolved and the presence of this filter is no longer necessary. Flattening filter free (FFF) beams exhibit higher dose rates, heterogeneous dose profiles, modified energy spectra and lower out-of-field dose. This PhD aimed at studying the characteristics of unflattened beams, as well as their impact in clinical utilization. Several subjects were thoroughly investigated: radiobiology, dosimetry, quality controls, modelling and treatment planning. In vitro experiments ensured that the high dose-rate of FFF beams had not a radiobiological impact. A wide review of the literature was conducted to corroborate these results. In order to understand thoroughly the characteristics of FFF beams, measurements were conducted using several detectors. The effect of the spectra and dose rates of unflattened beams on dose calibration were also studied. FFF beams were modeled in two TPSs. The methods, results and model parameters have been compared between the available beam qualities as well as between both TPSs. Furthermore, the implementation of stereotactic treatments technique was the occasion to investigate small beam dosimetry. Prostate cancer cases treated with VMAT and pulmonary tumors treated with stereotactic 3D beams were also studied. The comparison of dose distributions and treatment metrics give advantage to FFF beams. Mastering physical and biological aspects of flattening filter free beams allowed the IPC to start FFF treatments. Comparative studies have since resulted in a deeper understanding on the pertinent use of these beams.

No description available.

Arguments for miniaturization of a thermal field-flow fractionation system ( and #956;-ThFFF) and fabrication of a micro-scale acoustic field-flow fractionation system ( and #956;-AcFFF) using similar methods was presented. Motivation for miniaturization of ThFFF systems was established by examining the geometrical scaling of the fundamental ThFFF theory. Miniaturization of conventional macro-scale ThFFF systems was made possible through utilization of micromachining technologies. Fabrication of the and #956;-ThFFF system was discussed in detail. The and #956;-ThFFF system was characterized for plate height versus flow rate, single component polystyrene retention, and multi-component polystyrene separations. Retention, thermal diffusion coefficients, and maximum diameter-based selectivity values were extracted from separation data and found comparable with macro-scale ThFFF system results. Retention values ranged from 0.33 to 0.46. Thermal diffusion coefficients were between 3.0ױ0-8 and 5.4ױ0-8 cm2/sec?? The maximum diameter-based selectivity was 1.40. While the concept of an acoustic FFF sub-technique has been around for decades, the fabrication methods have not been available until recently. The theory was developed in full including relating sample physical properties to retention time in the FFF system. In addition to the theory, the design and fabrication of the and #956;-AcFFF was presented. Design results from an acoustic modeling program were presented with the determination of the acoustic resonant frequency. The acoustic-based systems was designed around the model results and characterized by electrical input impedance, fluidic, plate height, polystyrene suspension retention, and polystyrene mixture separation studies. The and #956;-AcFFF system was able to retain a series of nanometer scale polystyrene samples. However, the retention data did not follow normal mode retention but did reveal the location of the steric inversion point for the power level used, around 200 nm. The results of the multiple component separation confirmed this results as the sample, which contained 110, 210, and 300 nm diameter samples, was not resolved but only broadened.

Consumer additive manufacturing (3D printing) has rapidly grown over the last decade. While the technology for the most common type, Fused Filament Fabrication (FFF), has systematically improved and sales have increased, fundamentally, the capabilities of the machines have remained the same. FFF printers are still limited to depositing layers onto a flat build plate. This makes it difficult to combine consumer AM with other objects. While consumer AM promises to allow us to customize our world, the reality has fallen short. The ability to directly modify existing objects presents numerous possibilities to the consumer: personalization, adding functionality, improving functionality, repair, and novel multi-material manufacturing processes. Indeed, similar goals for industrial manufacturing drove the research and development of technologies like direct write and directed energy deposition which can deposit layers onto uneven surfaces. Replicating these capabilities on consumer 3-axis FFF machines is difficult mainly due to issues with reliability, repeatability, and quality. This thesis proposes, demonstrates, and tests a method for scanning and printing dimensionally-accurate (unwarped) digital forms onto physical objects using a modified consumer-grade 3D printer. It then provides an analysis of the machine design considerations and critical process parameters. / Master of Science / Consumer additive manufacturing (3D printing) has rapidly grown over the last decade. While the technology for the most common type, Fused Filament Fabrication (FFF), has systematically improved and sales have increased, fundamentally, the capabilities of the machines have remained the same. FFF printers are still limited to depositing layers onto a flat build plate. This makes it difficult to combine consumer AM with other objects. While consumer AM promises to allow us to customize our world, the reality has fallen short. The ability to directly modify existing objects presents numerous possibilities to the consumer: personalization, adding functionality, improving functionality, repair, and novel multi-material manufacturing processes. Indeed, similar goals for industrial manufacturing drove the research and development of technologies like direct write and directed energy deposition which can deposit layers onto uneven surfaces. Replicating these capabilities on consumer 3-axis FFF machines is difficult mainly due to issues with reliability, repeatability, and quality. This thesis proposes, demonstrates, and tests a method for scanning and printing dimensionally-accurate (unwarped) digital forms onto physical objects using a modified consumer-grade 3D printer. It then provides an analysis of the machine design considerations and critical process parameters.

Yes / There is an increased evidence for treating hypertension by a combination of two or more drugs. Increasing the number of daily intake of tablets has been reported to negatively affect the compliance of patients. Therefore, numerous fixed dose combinations (FDCs) have been introduced to the market. However, the inherent rigid nature of FDCs does not allow the titration of the dose of each single component for an individual patient's needs. In this work, flexible dose combinations of two anti-hypertensive drugs in a single bilayer tablet with a range of doses were fabricated using dual fused deposition modelling (FDM) 3D printer. Enalapril maleate (EM) and hydrochlorothiazide (HCT) loaded filaments were produced via hot-melt extrusion (HME). Computer software was utilised to design sets of oval bi-layer tablets of individualised doses. Thermal analysis and x-ray diffractometer (XRD) indicated that HCT remained crystalline in the polymeric matrix whilst EM appeared to be in an amorphous form. The interaction between anionic EM and cationic methacrylate polymer may have contributed to a drop in the glass transition temperature (Tg) of the filament and obviated the need for a plasticiser. Across all tablet sets, the methacrylate polymeric matrix provided immediate drug release profiles. This dynamic dosing system maintained the advantages of FDCs while providing a superior flexibility of dosing range, hence offering an optimal clinical solution to hypertension therapy in a patient-centric healthcare service.

The master's thesis deals with the determination of mechanical properties of materials used for 3D printing (PLA, PETG, ABS, ASA). The theoretical part contains characteristics of polymer materials and describes testing of mechanical properties and additive technologies. In the experimental part, the standardized samples produced by the technology Fused Filament Fabrication were analysed using mechanical tests (tensile test and hardness test). Some of the examples were exposed to the selected factors of degradation. Selected parameters from mechanical testing (tensile strength, modulus of elasticity in tension, elongation and Shore hardness) were statistically processed. The conclusion of the thesis evaluates the impact of the factors of degradation on particular materials and compares them subsequently.

Le but de la présente étude était de vérifier l'influence de deux compartiments du latex d'Hevea brasiliensis, les lutoïdes et le sérum C, sur le durcissement au stockage et sur la mésostructure du caoutchouc naturel (NR). L'implication des composants minéraux du latex a fait l'objet d'un focus spécial. La mésostructure du NR a été étudié par fractionnement par couplage flux-force à flux asymétrique couplé à un détecteur à diffusion de lumière multiangulaire (AF4-MALS) et par chromatographie d'exclusion de tailles équipée d'un détecteur de diffusion de lumière multiangulaire (SEC- MALS). La spectrométrie de masse couplée à une torche à plasma (ICP- MS) a été utilisée pour déterminer la composition en éléments minéraux du NR.L'AF4 - MALS et l'ICP-MS n'ayant jamais été utilisées pour l'analyse du NR, les méthodes ont été développées. Pour l'AF4 - MALS, la meilleure séparation entre les deux populations principales, chaînes de polyisoprène isolées (pelote statistique) et les microagrégats (Gel<1μ), a été obtenue avec une diminution linéaire, plutôt qu'exponentielle, du flux croisé. Pour l'ICP-MS, les optimisations réalisées concernent la quantité de NR à échantillonner, la méthodologie de solubilisation des cendres, la concentration des solutions de cendres et la gestion des interférences m/z . Tous les éléments, excepté le soufre, ont été analysés en utilisant un mélange H2/He comme gaz de collision-réaction (mode CCT H2/He). La teneur en soufre a été déterminée par le rapport m/z égal à 48 (32S16O+) en mode CCT O2.Les différents compartiments du latex des champs (crème, skim, sérum C et lutoïdes) ont été séparés par centrifugation à grande vitesse. L'évolution de la mésostructure des films obtenus à partir de ces trois latex; latex des champ (FL), le latex de crème (CL) et de latex de skim (SK), par un procédé de structuration lente (échantillons stockés à température ambiante dans le laboratoire pendant 3 mois) a été suivie par SEC- MALS. Le skim n'étant pas sensible au processus de structuration lente, le nombre des étapes de centrifugation a été réduit. La stabilité des lutoïdes a été étudiée par un paramètre qualitatif (état visuel des lutoïdes après centrifugation) et un paramètre quantitatif (indice d'éclatement ou BI). Bien que les deux méthodes n'aient pas donné de résultats strictement corrélées, le BI peut être un bon indicateur de la stabilité des lutoïdes. Pour les échantillons de FL, une bonne corrélation entre la stabilité des lutoïdes et le durcissement au stockage (P) a été observée. Pour déterminer si des composés du C-sérum étaient également impliqués dans le durcissement au stockage, des expériences supplémentaires ont été effectuées en ajoutant des quantités variables de sérum C ou de lutoïdes à des particules de caoutchouc purifiées. L'augmentation à la fois de les quantités de sérum C et des lutoïdes a entrainé une augmentation du durcissement au stockage (P).La mésostructure des films et des feuilles séchées de l'air (ADS) préparés à partir des FL et CL a été analysée par SEC-MALS et AF4-MALS. Pour les échantillons d'ADS, quelle que soit la technique utilisée, les échantillons de FL présentaient des Mw, Mn et Gel>1μ supérieurs aux échantillons de CL. Cette différence entre échantillons de FL et de CL n'a pas été observée pour les échantillons de films. / The aim of the present work was to study the influence of two Hevea brasiliensis latex compartments, namely lutoids and C-serum, on the storage hardening and on mesostructure of natural rubber (NR). A special focus was done on the involvement of mineral components of latex. The NR mesostructure was studied by asymmetrical flow field-flow fractionation coupled to a multiangular light scattering detector (AF4-MALS) and by size exclusion chromatography equipped with a multiangular light scattering detector (SEC-MALS). Inductively coupled plasma mass spectrometry (ICP-MS) was used to determine the mineral element composition of NR.As AF4-MALS and ICP-MS were never used for NR analysis, the methodologies were developed. For AF4-MALS, the best separation between the two main populations, namely isolated polyisoprene chains (random coil) and microaggregates (Gel<1µ) was given by a linear decrease, rather than exponential, of the cross-flow. For ICP-MS, the optimizations were in terms of amount of NR to be sampled, ash solubilisation methodology, ash solutions concentrations and m/z interference management. All elements, except sulfur, were determined using a mixture H2/He as collision-reaction gas (CCT H2/He mode). Sulfur content was determined through the m/z equal to 48 (32S16O+) in the CCT O2 mode.The different compartments of the whole field latex (cream, skim, C-serum and lutoids) were separated by high speed centrifugation. The mesostructure evolution of films obtained from these 3 lattices; whole field latex (FL), cream latex (CL), and skim latex (SK), by a slow structuring process (samples stored at room temperature in the laboratory for 3 months) was followed by SEC-MALS. As it was observed that the skim was not sensitive to the slow structuring, the centrifugation steps were reduced.Lutoid stability was studied by a qualitative parameter (visual lutoid status after centrifugation) and a quantitative parameter (bursting index or BI). Although the two methods could not provide strictly correlated results, BI can be a good indicator of lutoid stability. For the FL samples, a good correlation between the lutoid stability and storage hardening (ΔP) was observed. To determine if some compounds of C-serum are also involved in the storage hardening, additional experiments were done adding variable quantities of C-serum or lutoids to purified rubber particles. The storage hardening (ΔP) increased by the increase of both C-serum and lutoid quantities.The mesostructure of films and air dried sheet (ADS) made from FL and CL lattices (obtained from reduced centrifugation process) were analyzed by SEC-MALS and AF4-MALS. Concerning the ADS samples, whatever the technique used, FL samples exhibited a higher Mw, Mn, and Gel>1µ than CL samples. This difference between FL and CL samples was not observed for film samples. The microaggregates (Gel<1µ) were presented in all samples but the FL samples had more compact microaggregates, with a much higher Mw than the CL samples. Moreover, AF4 showed that the structure of microaggregates was very different between ADS and film samples. The Mw of microaggregates of ADS was 2 to 4 times higher than that of films. The mineral elements were determined only on samples from ADS (FL and CL). The main elements in NR were K, P, Mg, and S, in decreasing order. The purification of rubber particles affected the decrease in the element contents. During the storage of the latex at room temperature, only calcium content decreased, for both FL and CL samples.

Additive manufacturing has recently become an increasingly available, affordable, and versatile technology for industry and enthusiasts alike. One opportunity that this provides is using conductive filament to print electronic sensors and components. This thesis addresses some of the possibilities and limitations of using conductive filament for constructing electronic sensors and components. Using Protopasta Conductive polylactic acid filament, several components were printed and tested. Printed components include resistors, capacitors, strain gauges and touch sensors. To all components connector pad were added to create a uniform interface between filament and metal. Resistors were printed with varying geometry, ranging from lengths 12 mm to 60 mm and height 1 mm to 4 mm. Their resistances were measured and found to range from 250 Ω to 1600 Ω. Circular capacitors were printed with diameters, 2 cm, 3 cm and 4 cm respectively. However, the capacitors showed no capacitance during measurements, indicating that this material with its high resistivity is unsuitable for creating capacitors. Strain gauges and touch sensors were also constructed, showing that there is a potential for using this filament. The strain gauges had a resistance increase that depended heavily on geometry; straight components had a resistance increase up to 15 % for 40° angle bends, where as gauges with bends had close to zero increase. Two proof-of-concept prototypes of touch sensors were also constructed, showing that the material can be used for this purpose. Creating high quality prints with the filament proved difficult; it had a tendency of clogging the printer, printing unevenly, and showed poor layer adhesion. Primary future work is to improve the printing process. / Additiv tillverkning har de senaste decennierna alltmer blivit tillgängligt, billigare och mer mångsidigt, och kan nyttjas av industri och entusiaster. En möjlighet som uppkommit är att additiv tillverkning möjliggör konstruktion med konduktiva filament, för att printa exempelvis sensorer och elektriska komponenenter. Denna rapport behandlar möjligheter och begränsningar av att använda konduktivt filament för sensorer och elektriska komponenter. Med Protopasta Conductive PLA (PolyLactic Acid) filament har flera komponenter printats och testats. Printade komponenter inkluderar resistorer, kondensatorer, töjningsgivare och vidröringssensorer. Till alla komponent har tillagts kontaktytor för att skapa ett enhetligt gränssnitt mellan filament och metall. Resistorer har printats med varierande geometri, från längder mellan 12 mm och 60 mm till tjocklekar mellan 1 mm och 4 mm. Resistanserna mättes och låg mellan 250 Ω och 1600 Ω. Cirkulära kondensatorer printades med diametrarna 2 cm, 3 cm och 4 cm. Däremot visade kondensatorerna ingen kapacitans under mätningar, vilket indikerar att detta material med sin höga resistivitet är olämpligt för att skapa kondensatorer. Töjningsgivare och vidröringssensorer har också konstruerats. Töjningsgivaren har en resistans som beror kraftigt på geometrin; raka komponenter uppvisade en resistansökning på uppemot 15 % för 40° böjningar, medan töjningsgivare med öglor hade en resistansökning nära noll. Två prototyper av vidröringssensorer konstruerades och verifierades, vilket visar att materialet kan användas till detta syfte. Att få högkvalitativa utskrifter med filamentet visade sig vara svårt; filamentet tenderade till att täppa till skrivarens munstycke, ge ojämna utskrifter och uppvisade dålig fästning mellan lagren. Huvudsakligt framtida arbete är att förbättra utskriftsprocessen.

Denna rapport detaljerar ett examensarbete där användbarheten på en lokalt skapad 3D-skrivare förbättrats i syfte av att tillåta skrivarens användning av studenter på KTH Södertälje. Skrivarens har ursprungligen byggts för att agera som ett pågående projekt som kan återupptas som examensarbete, varav detta är det tredje. Skrivaren i fråga är en FFF-skrivare i ”box” stil med utskriftsvolymen 400 x 400 x 375 mm. Vid projektstart är skrivaren i nominellt användbart skick men saknar dokumentation, processrutiner och användbara slicerprofiler. Skrivaren besitter även hårdvaran för att tillåta utskrifter med två material men denna funktion har aldrig implementerats. Framtagningen av slicerprofiler har tagit upp majoriteten av arbetet och har utförts genom iterativ testning där inställningar modifieras och testas, vartefter positiva ändringar implementeras för att gradvis eliminera defekter och minska haveririsk. Projektet har levererat en full slicerprofil som tillåter pålitliga utskrifter med två olika PLA-filament samt en användarmanual till skrivaren. Utöver användarverktygen så detaljeras även de reparationsarbeten och hårdvaruändringar som gjorts. Profil, manual, rådata och filer för utskrivna komponenter är tillgängliga i en projektmapp som kan hittas från en länk i början av kapitel 4. / This report details a thesis project for getting a 3D-printer internally built at KTH Södertälje into a state where it can be used by other students for unrelated projects. The printer acts as an ongoing project that can be used for thesis work, with this project in particular being the third in line. The printer in question is an FFF-printer with the ”box” configuration and a print volume of 400 x 400 x 375 mm. At project start, the printer is nominally in working condition but lacks any documentation for proper handling and slicer settings. Additionally, its multi-material function has never been successfully implemented. Adjustment of slicer settings is done using iterative testing, wherein settings are modified and tested, with positive changes being integrated and negative changes discarded with the intent of removing defects and minimizing risk of print failure. The project has delivered a slicer profile for Cura capable of reliably printing with two different PLA filaments as well as a user manual. Hardware upgrades and repair work has also been detailed. The profile, user manual, raw data and object files for upgrades and spares can be found through a link in the beginning of chapter 4.