Search

Silica nanoparticles (SNPs) are one of the most widely used and studied nanoparticles. Their advantages include simple synthesis, biocompatibility, stability, simple chemical or biochemical surface modification. Many scientific studies reported optimization of conditions for size controlled synthesis of SNPs, yet laboratory scale production of monodisperse SNPs with desired size in the range of 3–200 nm remains challenging. In this work we have performed a systematic examination of the influence of 4 main parameters on the size in synthesis SNPs ; temperature, concentration of tetraethoxysilane (TEOS), ammonia and water. The synthesis was performed based on the work of Stöber, Fink and Bohn. For all experiments we have used vials with the same size, same magnetic stirring bars, constant stirring rate and ethanol as a solvent. TEOS was freshly distilled and the concentration of ammonium hydroxide was determined with titration, prior to all experiments. The series of experiments was performed by changing only one of the parameters. The concentration of TEOS was varied from 0.01-0.26 M, ammonia as a catalyst 0.097-0.291 M and water 2-5 M. Temperatures were in the range of 25-55°C. At low TEOS concentrations there was a very low yield of SNPs, for this reason the concentration was set to 0.26 M for all other experiments. A liner correlation between the size and temperature was found, as well as for size and water concentration. This liner correlation remained for three different ammonia concentrations, but different SNPs size ranges. After statistical analysis of experimental data we have found that the size of the SNPs can be engineered in the 5-200 nm range with good reproducibility and confidence interval by changing one of the main parameters of synthesis. Best results can be obtained by altering the concentration of ammonia or temperature.

Software systems have become ubiquitous in our everyday life. They replace humans for critical tasks that involve high costs and even human lives. The serious consequences caused by the failure of such systems make crucial the use of rigorous methods for system validation. One of the widely-used formal verification methods is the automata-theoretic approach to model checking. It takes as input a model of the system and a property, and answers if the model satisfies or not the property. To achieve this goal, it translates the negation of the property in an automaton and checks whether the product of the model and this automaton is empty. Although it is automatic, this approach suffers from the combinatorial explosion of the resulting product. To tackle this problem, especially when checking stutter-invariant LTL properties, we firstly improve the two-pass verification algorithm of Testing automata (TA), then we propose a transformation of TA into a normal form (STA) that only requires a single-pass verification algorithm. We also propose a new type of automata: the TGTA. These automata also enable a check in a single-pass and without adding artificial states : it combines the benefits of TA and generalized Büchi automata (TGBA). TGTA improve the explicit and symbolic model checking approaches. In particular, by combining TGTA with the saturation technique, the performances of the symbolic approach has been improved by an order of magnitude compared to TGBA. Used in hybrid approaches TGTA prove complementary to TGBA. All the contributions of this work have been implemented in SPOT and LTS-ITS, respectively, an explicit and a symbolic open source model-checking libraries.; Les systèmes logiciels sont devenus omniprésents se substituant à l'homme pour des tâches délicates, souvent critiques, mettant en jeu des coûts importants voire des vies humaines. Les conséquences des défaillances imposent la recherche de méthodes rigoureuses pour la validation. L'approche par automates du model-checking est la plus classique des approches de vérification automatique. Elle prend en entrée un modèle du système et une propriété, et permet de savoir si cette dernière est vérifiée. Pour cela un model-checker traduit la négation de la propriété en un automate et vérifie si le produit du système et de cet automate est vide. Hélas, bien qu'automatique, cette approche souffre d'une explosion combinatoire du nombre d'états du produit.Afin de combattre ce problème, en particulier lors de la vérification des propriétés insensibles au bégaiement, nous proposons la première évaluation d'automates testeur (TA) sur des modèles réalistes, une amélioration de l'algorithme de vérification pour ces automates et une méthode permettant de transformer un TA en un automate (STA) permettant une vérification en une seule passe.Nous proposons aussi une nouvelle classe d'automates: les TGTA. Ces automates permettent une vérification en une seule passe sans ajouter d'états artificiels. Cette classe combine les avantages des TA et des TGBA (automates de Büchi). Les TGTA permettent d'améliorer les approches explicite et symbolique de model-checking. Notamment, en combinant les TGTA avec la saturation, les performances de l'approche symbolique sont améliorées d'un ordre de grandeur par rapport aux TGBA.

International audience; Geldenhuys and Hansen showed that a kind of ω-automata known as Testing Automata (TA) can, in the case of stuttering-insensitive properties, outperform the Büchi automata traditionally used in the automata-theoretic approach to model checking. In previous work, we compared TA against Transition-based Generalized Büchi Automata (TGBA), and concluded that TA were more interesting when counterexamples were expected, otherwise TGBA were more efficient. In this work we introduce a new kind of automata, dubbed Transition-based Generalized Testing Automata (TGTA), that combine ideas from TA and TGBA. Implementation and experimentation of TGTA show that they outperform other approaches in most of the cases.