The ubiquitous presence and chronic effect of pharmaceuticals is one of the emerging issues in environmental field. As a result of incomplete removal by sewage treatment plants, pharmaceuticals are released into the environment and drinking water sources. On the other hand, conventional drinking water treatment processes such as coagulation, filtration and sedimentation are reported to be ineffective at removing pharmaceuticals. Therefore, the potential presence of pharmaceuticals in finished drinking water poses a threat on public health. Antibiotics, as an important group of pharmaceuticals, are given special concerns because the potential development of bacteria-resistance. Ozonation and advanced oxidation processes are demonstrated to be quite effective at removing pharmaceuticals. The oxidation of pharmaceuticals is caused by ozone itself and hydroxyl radicals that are generated from ozone decomposition. Whether ozone or hydroxyl radicals are the primary oxidant depends on the specific pharmaceutical of interest and the background water matrix. In this research, erythromycin, a macrolide antibiotic, was chosen as the target compound because of its high detection frequency in the environment and its regulation status. The objective of this research was to investigate the removal performance of erythromycin by ozonation from the standpoint of kinetics. The effects of pH, carbonate and phosphate buffers, and initial ozone dose on ozonation of erythromycin were also studied. The second-order rate constant for the reaction between deprotonated erythromycin and ozone was determined to be 4.44x10⁹ M⁻¹·s⁻¹ while protonated erythromycin did not react with ozone. Ozone was determined to be the primary oxidant for erythromycin removal by ozonation. pH was found to have great positive impact on the degradation of erythromycin by ozonation due to the deprotonation of erythromycin at high pH. Carbonate and phosphate buffers were found to have negligible effects on the degradation of erythromycin by ozonation. Initial ozone dose showed a positive impact on the total erythromycin removal rate by ozonation. / text

The resistance random access memory (RRAM) is one of the most popular of the next generation memories with the high operating speed, reliability and the smallest miniature size. RRAM has metal-insulator-metal structure that can greatly reduce the difficulty of entry, but the biggest problem is how to choose the insulator. We selected silicon-based materials to match the intergrated circuits manufacturing process. In this work, sputtering titanium doping in the silicon oxide thin film has a stable characteristic of resistance switching. By material analyzing, we found that supercritical carbon dioxide fluid (SCCO2) treatment can passivate the silicon oxide defect and the self-reduction of titanium oxide, but it also brought OH group into our thin film. So we observed the interface type characteristic of resistance switching. Using constant voltage sampling experiment extract the reaction rate constant (k) and the active energy, prove that the reaction is caused by OH injection. Double-layer structure with titanium-doped and carbon-doped silicon oxide RRAM promote lower operating current by hopping conduction, which is caused by graphite oxide doping. The Space-Charge Limited Current mechanism for high limited current is proven by COMSOL electric field simulation.

Les espèces réactives de l'oxygène sont générées dans l'environnement biologique dans le cadre du métabolisme, mais elles peuvent aussi être produites en excès dans le cas de stress oxydatif provoqué par exemple par une exposition aux rayons UV. Dans le travail présenté ici, nous sommes intéressés par l'oxydation des protéines par deux de ces espèces réactives de l'oxygène : le peroxyde d'hydrogène, oxydant plutôt faible avec un temps de vie long, et l'oxygène singulet, oxydant fort avec un temps de vie court. L'action de ce dernier sur les protéines est étudiée en utilisant la spectroscopie de phosphorescence résolue en temps et l'oxydation des protéines par le peroxyde d'hydrogène est suivie par spectroscopie Raman. Dans ce cas, un travail préliminaire a été nécessaire afin d'attribuer de manière précise les bandes Raman des chaînes latérales des résidus d'acides aminés. Pour les deux types d'oxydations, les constantes de vitesse des réactions ont été déterminées pour trois protéines modèles. La stratégie suivie est d'utiliser de petits fragments de protéines tels que des acides aminés libres et des tripeptides pour comprendre ce qui se passe à l'échelle de la protéine. Cela nous aide à souligner l'importance de l'environnement protéique. Dans le cas de l'étude par spectroscopie Raman, l'influence du nombre de liaisons peptidiques sur les spectres obtenus depuis l'acide aminé libre, au tripeptide, jusqu'à la protéine est aussi mis en évidence. / Reactive species of oxygen are generated in biological environment as part of metabolism but they can also be produced in excess in case of oxidative stress provoked by UV exposure for example. In the present work we are interested in the oxidation of proteins by two of those reactive species of oxygen : hydrogen peroxide, rather weak oxidant with a long life time, and singlet oxygen, a strong oxidant with a short life time. The action of the latter one on the proteins is studied by using time-resolved phosphorescence spectroscopy and oxidation of the proteins by hydrogen peroxide is monitored by using Raman spectroscopy. In this case a preliminary work was necessary to attribute accurately the Raman bands of amino-acid residues side chains. For both oxidations, reaction rate constants of the reactions were determined for three model proteins. The strategy followed is to use small fragments of proteins such as free amino-acids and tripeptides to understand what is happening at the protein scale. This helps us underlining the importance of the proteic environment. In the case of Raman spectroscopy study, it also shows the influence of the number of peptidic bonds on the spectra obtained from free amino-acid to tripeptide and then to protein.Keywords : Raman Spectroscopy, protein, oxidation, hydrogen peroxide, singlet oxygen, time-resolved phosphorescence spectroscopy, reaction rate constant.

This work aims at modelling and evaluating a new type of photocatalytic reactors, named fixed-bed photocatalytic membrane reactor (FPMR). Such reactors are based on the deposition of a thin layer of photocatalysts on a permeable substrate by filtration. This layer serves as a photocatalytic membrane, named fixed-bed photocatalytic mem-brane (FPM), which is perpendicularly passed by the reactant solution and illuminated by a suitable light source. One advantage of FPMs is their renewability. The model, which was developed for this reactor, relates the overall reaction rate in the FPM with the intrinsic reaction kinetic at the catalyst surface and accounts for light intensity, structural and optical layer properties as well as the mass transfer in the pores. The concept of FPMR was realised by using a flat sheet membrane cell. It facilitated principal investigations into the reactor performance and the validity of the model. For this purpose, the photocatalytic degradation of organic compounds, such as meth-ylene blue and diclofenac sodium, was conducted at varying conditions. Pyrogenic ti-tania was used as a photocatalyst. The experimental data support the developed mod-el. They also indicate a significant impact of the flow conditions on the overall photo-catalytic activity, even though the Reynolds number in the FPM was very small; the to-tal mass transfer rate in the FPM amounted to more than 1.0 s−1. The experiments also showed a sufficient structural strength of the FPM and photocatalytic stability. In addi-tion, the renewal and regeneration of FPMs was successfully demonstrated. Furthermore, another FPMR was designed by means of submerged ceramic mem-branes. This reactor was mainly used to assess the effectiveness and efficiency of FPMRs at the example of the photocatalytic degradation of oxalic acid. The correspond-ing reactor was run closed loop and in continuous mode. The effectiveness of the reac-tor was evaluated based on common descriptors, such as apparent quantum yield, photocatalytic space-time yield and light energy consumption. The results showed that the FPMR based on submerged ceramic membrane had a higher efficiency than other reported photocatalytic reactors. The comparison of the different modes of operation revealed that the closed loop FPMR is most efficient with regard to light energy con-sumption. Finally, this work discusses the up-scaling of FPMRs for industrial applications and proposes a solution, which can e.g. be employed for wastewater treatment or CO2 conversion.:Abstract iii Kurzfassung v Acknowledgment vii Contents ix Nomenclature xiii 1 Introduction 1 1.1 Motivation 1 1.2 Aim and objectives of the work 3 1.3 Thesis outline 3 2 Heterogeneous photocatalytic reactors 5 2.1 Introduction to photocatalysis 5 2.2 Processes in heterogeneous photocatalysis 6 2.2.1 Optical phenomena 7 2.2.2 Mass transfer 8 2.2.3 Adsorption and desorption 9 2.2.4 Photocatalytic reactions 10 2.2.5 Factors affecting heterogeneous photocatalysis 12 2.3 Photocatalytic reactor systems towards water treatment 16 2.3.1 Introduction to photocatalytic reactors 16 2.3.2 Development of photocatalytic reactor designs 17 2.3.3 Quantitative criteria for evaluating photocatalytic reactor designs 21 2.4 Cake layer formation in membrane microfiltration 22 2.4.1 Suspension preparation 22 2.4.2 Cake layer formation 23 2.5 Fluid flow through a fixed bed of particles 25 2.5.1 Pressure drop through a fixed-bed 25 2.5.2 Liquid-solid mass transfer correlation in fixed-bed 25 3 Concept and mathematical modelling of FPMRs 29 3.1 Concept of fixed-bed photocatalytic membrane reactors 29 3.2 Modelling of fixed-bed photocatalytic membrane reactors 31 3.3 Model sensitivity analysis 37 3.4 Chapter summary 39 4 FPMR realised with flat sheet polymeric membrane 41 4.1 Introduction 41 4.2 Materials and set-up 41 4.2.1 Materials 41 4.2.2 Experimental set-up 43 4.3 Experiments and methods 48 4.3.1 Formation of fixed-bed photocatalytic membrane 48 4.3.2 Reactor performance 50 4.3.3 Parameters study and model verification 53 4.3.4 Catalyst layer characterisation 56 4.3.5 Measurement and evaluation of photocatalytic activity of FPM 59 4.4 Results and model verification 60 4.4.1 Reactor performance 60 4.4.2 Influence parameters 71 4.4.3 Model verification 79 5 FPMR realised with submerged ceramic membrane 92 5.1 Introduction 92 5.2 Materials and reactor set-up 93 5.2.1 Reactor set-up 93 5.2.2 Chemicals 97 5.3 Experiments and methods 97 5.3.1 Formation of fixed-bed photocatalytic membranes 97 5.3.2 Photocatalytic performance 97 5.3.3 Parameter study 98 5.3.4 Reactor model for calculating reaction rate constant of FPM 99 5.3.5 Comparison of different reactor schemes 102 5.4 Results and discussions 105 5.4.1 Reactor performance 105 5.4.2 Consistency of CPMR and LPMR data 107 5.4.3 Influence of catalyst loading 108 5.4.4 Influence of permeate flux and light intensity 109 5.4.5 Reactor efficiency 111 5.4.6 Comparison of different reactor schemes 113 5.5 Proposed up-scaled FPMR systems 113 5.6 Concluding remarks 116 6 Conclusion and outlook 118 6.1 Summary of thesis contributions 118 6.2 Discussion and outlook 120 References 122 List of Figures 134 List of Tables 138 Appendix A Calibration 139 A.1 Distribution of light intensity on the surface of catalyst layer 139 A.2 Concentration and absorbance of diclofenac 141 A.3 TOC concentration and electrical conductivity of oxalic acid 141 A.4 Concentration and absorbance of methylene blue 142 Appendix B Mathematical modelling 143 B.1 Influence of axial dispersion on the reaction rate 143 B.2 Special case 146 Appendix C Comparison the photocatalytic activity of TiO2 and ZnO 147 Appendix D Mathematical validation of model for LPMR and CPMR 148 D.1 Model for LPMR (cf. Eq. (5 12)):148 D.2 Model for CPMR (cf. Eq. (5 17)) 149 Appendix E Particle size distribution 151