Radiation chemistry is concerned with understanding the chemical kinetics following the application of ionising radiation. There are two main methods for modelling recom- bination and spin dynamics in radiation chemical systems: The Monte Carlo random flights algorithm, in which the trajectories of the diffusing species are followed ex- plicitly and the Independent Reaction Times (IRT) algorithm, where reaction times are sampled from appropriate marginal distribution functions. This thesis reports develop- ments to both methods, and applies them to better understand experimental findings, particularly spin relaxation effects. Chapter 4 introduces current simulation techniques and presents newly developed algorithms and simulation programs (namely Hybrid and Slice) for modelling spatially dependent spin effects. A new analytical approximation for accurately treating ion-pair recombination in low-permittivity solvents in also presented in this chapter. Chapter 5 explores the photodissociation of H₂O₂, where there is some controversy in the literature on the spin state of the precursor. This chapter explores the possibility of reproducing the observed spin polarisation phase using the Radical Pair Mechanism. Chapter 6 presents two new algorithms for treating reactive products in the IRT framework. These have been tested for two chemical systems: (i) photodissociation of H₂O₂ where the ·OH are scavengeable; (ii) water photolysis which produces H⁺, ·OH and e⁻__aq. In the latter case a careful handling of three body correlations is required. Chapter 7 presents simulation results which suggest a strong correlation between scavenging and ion recombination in low permittivity solvents. A path decomposition method has been devised that allows IRT simulations to be corrected for this effect. Chapter 8 presents evidence for spin-entanglement and cross-recombination to act as an extra source of relaxation for ion-recombination in low permittivity solvents. It is hypothesised this effect contributes to the anomalous relaxation times observed for certain cyclic hydrocarbons. Chapter 9 presents an extension of the IRT simulation method to micelles. The kinetics are shown to be accurately described using the mean reaction time and the exponential approximation.

Protein adsorption at the solid/liquid interface is of wide ranging importance in many different areas of science such as biomaterial design, the fate of nanoparticles and in the food industry. As a result, many studies have been undertaken with varying foci but there still remains a lack of agreement between many working in this field and fundamental questions regarding the adsorption of proteins at the solid/liquid interface. Neutron reflectivity is a powerful technique for probing the properties of adsorbed layers at interfaces due to its high structural resolution and the possibility of using isotopic substitution to distinguish between components of a mixture. In this work, neutron reflectivity has been used as the primary technique for the investigation of proteins adsorbed sequentially or from a binary mixture. Initially, the adsorption of four proteins (carbonic anhydrase II, lysozyme, human serum albumin and maltose binding protein) onto a clean silica surface was investigated which revealed the importance of electrostatic interactions and entropic contributions to the driving forces for adsorption. Most of the adsorbed layers were described by a 2-layer model with a thinner, denser layer adjacent to the surface and a thick, diffuse layer extending into the bulk solution. The presence of impurities is also shown to have a significant impact on the adsorption of HSA. A study of the HSA/myristic acid system shows that the presence of small amphiphiles can inhibit HSA adsorption and also remove a pre-adsorbed layer. A comparison was made between the protonated and deuterated forms of two proteins, HSA & MBP, showing the deuterated proteins to have a higher affinity for the surface with adsorption occurring in a 3:1 ratio when from a 1:1 mixture. Likewise, d-MBP displaced h-MBP more readily than vice versa in an investigation into the effect of incubation time on the properties of the protein layer. The extent of desorption into protein free buffer is not affected by incubation time but the extent to which d-MBP was displaced by h-MBP showed a clear trend of decreased exchange with increasing incubation time indicating an active exchange process was occurring. This was also observed to a lesser extent for the sequential adsorption of binary protein systems, HSA & LYS and HSA & MBP. When investigating binary protein mixtures the higher propensity for deuterated proteins to adsorb is observed. LYS dominates when adsorbed from a mixture with h-HSA but from a d-HSA & LYS mix both proteins were adsorbed. The marked difference between the adsorption characteristics of perdeuterated proteins and their protonated counterparts provides a good case study for testing the neutron reflectivity technique when investigating systems with more than one component. This thesis assesses the limitations of the methodology of contrast variation for investigating mixtures as well as using different solvent contrasts. A comparison of neutron reflectivity and dual polarisation interferometry (DPI) shows that the two techniques are similar and any small differences can be attributed to the small change in surface chemistry. This comparison also highlights the advantages of DPI; high throughput of samples and detailed information but the restriction to using a 1-layer model limits its use.

This thesis aims to investigate the possibility of developing spectroscopic techniques for trace gas detection, with particular emphasis on their applicability to breath analysis and medical diagnostics. Whilst key breath molecules such as methane and carbon dioxide will feature throughout this work, the focus of the research is on the detection of breath acetone, a molecule strongly linked with the diabetic condition. Preliminary studies into the suitability of cavity enhanced absorption spectroscopy (CEAS) for the analysis of breath are carried out on methane, a molecule found in varying quantities in breath depending on whether the subject is a methane-producer or not. A telecommunications near-infrared semiconductor diode laser (1.6 µm) is used with an optical cavity based detection system to probe transitions within the vibrational overtone of methane. Achieving a minimum detectable sensitivity of 600 ppb, the device is used to analyse the breath of 48 volunteers, identifying approximately one in three as methane producers. Following this, a second type of laser source, the novel and widely tunable Digital Supermode Distributed Bragg Reflector (DS-DBR) laser, is characterised and the first demonstration of its use in spectroscopy documented. Particular emphasis is given to its application to CEAS and to probing the transitions of the two Fermi resonance components of the CO_2 3ν_1 + ν_3 combination bands found within the spectral range (1.56 - 1.61 µm) of the laser, providing the means to determine accurate ^{13}CO_2/^{12}CO_2 ratios for use in the urea breath test. Not all molecules exhibit narrow, well-resolved ro-vibrational transitions and the next section of the thesis focuses on the detection of molecules, such as acetone, with broad, congested absorption features which are not readily discernible using narrowband laser sources. To provide the necessary specificity for these molecules, two types of broadband source, a Superluminescent Light Emitting Diode (SLED) and a Supercontinuum source (SC), both emitting over the 1.6 - 1.7 µm region, are used in the development of a series of broadband cavity enhanced absorption (BB-CEAS) spectrometers. The three broadband absorbers investigated here, butadiene, acetone and isoprene, all exhibit overtone and combination bands in this spectral region and direct absorption measurements are taken to determine absorption cross-sections for all three molecules. The first BB-CEAS spectrometer couples the SLED device with a dispersive monochromator, attaining a minimum detectable sensitivity of 6 x 10^{-8} cm^{-1}, which is further enhanced to 1.5 x 10^{-8} cm^{-1} on replacing the monochromator with a Fourier Transform interferometer. The spectral coverage is then extended to 1.5 - 1.7 µm by coupling the first SLED with a second device, providing a demonstration of simultaneous multiple species detection. Finally, a SC source is used to provide greater power and uniform spectral intensity, resulting in an improved minimum detectable sensitivity of 5 x 10^{-9} cm^{-1}, or 200 ppb, 400 ppb and 200 ppb for butadiene, acetone and isoprene respectively. This device is then applied to acetone-enriched breath samples; the resulting spectra are fitted with a simulation to return the acetone levels present in the breath-matrix. Following this, the development of a prototype breath acetone analyser, carried out at Oxford Medical Diagnostics Ltd. (OMD), is described. To fulfill the requirements of a compact and commercially-viable device, a diode laser-based system is used, which necessitates a thorough investigation into all possible sources of absorption level change. Most notably, this includes a study into the removal and negating of interfering species, such as water vapour, and to a lesser extent, methane. A novel solution is presented, utilising a water-removal device in conjunction with molecular sieve so that each breath sample generates its own background, which has allowed breath acetone levels to be measured within an uncertainty of 200 ppb. Spectroscopic detection then moves to the mid-infrared with the demonstration of a continuous wave 8 µm quantum cascade laser, which allows the larger absorption cross-sections associated with fundamental vibrational modes to be probed. Following the laser's characterisation using methane, including a wavelength modulation spectroscopy study, the low effective laser linewidth is utilised to resolve rotational structure in low pressure samples of pure acetone. Absorption cross-sections are determined before the sensitivity of the system is enhanced for the detection of dilute concentrations of acetone using two types of multipass cells, firstly a White cell and secondly a home-built Herriott cell. This allows an acetone minimum detectable absorption of 350 ppb and 20 ppb to be attained, respectively. Following this, an optical cavity is constructed and, on treating breath samples in a water-removal device prior to analysis, breath acetone levels determined and corroborated with a mass spectrometer. Finally, a preliminary study probing acetone in the ultraviolet is presented. Utilising an LED centred at 280 nm with a low finesse optical cavity and an imaging spectrograph, detection of 25 ppm of acetone is demonstrated and possible vibronic structure resolved. Combining large absorption cross-sections with the potential to be compact and commercially viable, further development of this arrangement could ultimately represent the optimum solution for breath acetone detection.

This thesis concerns several advances in electroanalytical chemistry which are separated into four parts: the electrochemical investigation of diffusional behaviour, the mechanistic and kinetic study of electrochemistry with room temperature ionic liquids (RTILs), the study of weakly-supported electrochemistry and a comparison of the Butler-Volmer and Marcus-Hush kinetic theories of electron transfer. A study of the diffusional behaviour of electroactive species is essential for further studies, especially in the case when electrochemistry is complicated through ion-pairing interactions between the electroactive species and other electrolytes. In Part II of this thesis, the possibility of the ferricenium ion-paired with perchlorate and hexauorophosphate in acetonitrile was discussed firstly employing chronoamperometric technique. Afterwards, the hexaammineruthenium III/II couple supported by chloride, nitrate and sulfate respectively was studied by a similar method. In order to avoid unwanted ion-pairing effects, room temperature ionic liquids can be applied as solvent, which provide high conductivity by their own ionic nature so that experiments can be conducted without adding additional supporting ions. Because of RTILs have distinctive properties, for example, high viscosity, high conductivity and ionic nature, electrochemistry could be greatly changed compared to those in conventional solvents. Part III of this thesis gives a detailed description of this topic. First, a study of the reduction of 1,4-benzonquinone in 1-ethyl-3-methylimidazolium bis(triuoromethanesulfonyl)imide is presented to show the new mechanistic insight into comproportionation in a electrochemical process. Second, a discussion of the oxidation of hydroquinone in the same RTIL is introduced to suggest a possible ECE scheme which was never reported before. The interest of weakly supported electrochemistry is also well-established, which not only provides another alternative strategy to avert ion-pairings but also offers more physical insights into electrochemical processes. Quantitative methods analysing voltammetries without an excess amount of supporting electrolyte are developed by introducing a migration term into the mass transport equation. In Part IV, new mechanistic insights into the reduction of 2-nitrobromobenzene and the dimerisation of 2,6-diphenylpyrylium in acetonitrile were provide by using weakly-supported cyclic voltammetry. Also, pulse techniques was also adopted to investigate the reduction of cobaltocenium and cobalt(III) sepulchrate, giving an alternative way for electrochemical analysis. A major application of electroanalytical chemistry is investigating electrochemical kinetics. Two kinetic models mostly concerned by electrochemists are Butler-Volmer and Marcus-Hush formalisms. The classic phenomenological model, Butler-Volmer formalism successfully describes most common electron transfer kinetics but shows little reference with nature of the involved species, solution and electrode material, while a more physically insightful theory, the Marcus-Hush formalism, takes species natural properties, for instance, a change of distances or geometry in the solvation or coordination shells of the redox, into account although it requires more complex formulations. Comparative studies of these two theories are presented in Part V in order to improve our understanding of the electron transfer kinetics under different circumstances. First, comparison of cyclic voltammograms of the reduction of europium(III) and 2-methyl-2-nitropropane at mercury microhemispherical electrodes was carried out. Second, square wave and differential pulse voltammetric techniques were also employed to further discriminate the two kinetic models. These studies all find that the symetric Marcus-Hush theory assuming the reactants and products have identical force constant dose not satisfactorily agree with the experimental results. Hence, the introduction of asymmetric Marcus-Hush theory was presented considering different oxidative and reductive reorganization energies, which gives reasonable agreement with experiments and makes this theory more insightful.

Graphene, as a recently discovered carbon allotrope, possesses with it many outstanding properties ranging from high electrical conductivity to great mechanical strength. Single layer graphene can be prepared by mechanical cleavage of graphite or by a more sophisticated method, CVD. However, the scale-up process for these preparation techniques is still unconvincing. Solution-processed graphene from exfoliation of graphite oxide on the other hand provides an alternative prospect resulting in the formation of graphene nanoplatelets (GNPs), which can be readily manipulated to tailor-suit various application demands. The main aim of the thesis is to explore the possibility and availability of this versatile method to produce graphene nanoplatelet and its composites with good all-round performance in energy and bioanalytical applications. A range of physical and chemical characterisation techniques were utilised including SEM, TEM, AFM, XPS, XRD, DLS, FTIR, Raman and UV-Vis spectroscopy in order to investigate the structural and chemical information of the graphene-based materials prepared. Functionalisation of graphene oxide with polyelectrolyte polymer could facilitate deposition of platinum nanoparticles in the formation of Pt-GNPs composites. The resultant composite was employed for bioanalytical application in the detection of an important neurotransmitter, glutamate, based on glutamate oxidase enzyme. The performance of Pt-GNPs based glutamate sensor exhibited enhanced sensitivity and prolonged stability compared to the sensors based on Pt decorated diamond or glassy carbon electrodes. The significant interfering effect from concomitant electrochemically active biological compounds associated with Pt-GNPs electrode however could be alleviated via opting for Prussian blue deposited GNPs electrode instead. The oppositely charged Pt-GNPs due to different functionalising polymers were also subject to self-assembly, which was enabled by the electrostatic interaction of the opposite charges of Pt-GNPs. The self-assembled film showed enhanced mechanical stability than the conventional drop-casted film and provided reasonably good activity towards oxidation of hydrogen peroxide. Three-component composite of graphene, nanodiamond and polyaniline was prepared via in-situ polymerisation for usage as an electrode material in electrochemical capacitors ("supercapacitors"). The addition of graphene was shown to significantly enhance specific capacitance while nanodiamond could improve the stability of the electrode by strengthening the polymer core. Another approach to produce a supercapacitor was via electrodeposition of nickel and cobalt hydroxides on graphene oxide film corporated with bicarbonate salt. The film was then subject to thermal reduction of GO and expansion of graphene layers within the film was observed. This leavening process enhanced the surface area of graphene film and thus the higher specific capacitance was obtained. The decoration of nickel and cobalt hydroxides onto the film also boosted the specific capacitance further however the poor cycling stability of the heated film still remained an issue. Graphene nanoplatelets were also used as a support for electrodeposition of Pt nanoparticles for methanol oxidation in acidic media. The preferential phase of the Pt deposited and large surface area of graphene in comparison to other carbon supports studied led to good catalytic activity being observed.

Parâmetros teóricos obtidos pela otimização de geometria da estrutura molecular em fase gasosa com o método quântico semi-empírico PM3, foram usados para correlacionar constantes de partição (log Poct; log Pc16 e log Pcic) para uma série de 700 solutos, considerando como referência as LSER obtidas com os parâmetros experimentais dos solutos: acidez (Σα2), basicidade (Σβ2), e dipolaridade/polarizabilidade (π2) de ligação de hidrogênio; refração molar de excesso (R2) e volume característico (Vx). Os parâmetros avaliados foram: as energias HOMO e LUMO; o momento de dipolo elétrico e as cargas atômicas parciais do hidrogênio mais positivo e do átomo mais negativo (classe IV - modelo CM1 ); e o índice de polarizabilidade como proposto por Famini e col.. Estes parâmetros teóricos, juntamente com Vx, são adequadas para serem utilizados em LSER na correlação dos valores de log P. Para a maioria dos solutos considerados, as quantidades eletrostáticas q+ , Ιq-Ι e µ descrevem adequadamente os termos de acidez, basicidade e dipolaridade de ligação de hidrogênio, o que não ocorre para solutos com mais de um polo positivo ou negativo relevante, como hidroquinona, nem para solutos como 2-nitrofenol com capacidade de formar ligação de hidrogênio intramolecular. As LSER para log Poct; log Pc16 e log Pcic com estes parâmetros teóricos são quimicamente consistentes. Cargas atômicas parciais, largamente usadas qualitativamente para descrever a reatividade de compostos e de grupos funcionais, podem ser usadas como descritor quantitativo de interações moleculares em estudos de solubilização, o primeiro estágio, que governa muitas reações químicas. / Semiempirical quantum theoretical parameters, obtained from PM3 geometry optimization method in gas-phase for a wide range of 700 solutes, were used to correlate experimental partition coeficients (log Poct; log Pc16 and log Pcyc), bearing as reference the LSER with experimentally derived parameters: Hydrogen bond acidity (Σα2), basicity (Σβ2) , and dipolarity/polarizability (π2); excess molar refraction and characteristic volume (Vx). Theoretical quantities tested includes: HOMO and LUMO energies; class IV, CM1 model atomic partial charges (q+ and Ιq-Ι) and dipole moment (µ); Famini\'s polarizability index (πF); and Vx. Those quantum derived quantities can successfully be used, jointly with Vx, to correlate log P values in the LSER approach. For most of the considered solutes, electrostatic quantities q+, Ιq-Ι and µ describes fairly well Hydrogen bond acidity, basicity and dipolarity terms, but do not describe hydrogen bond terms for solutes with more than one relevant positive or negative centers, like hydroquinone, nor solutes like 2-nitrophenol wich forms intramolecular hydrogen bond when in condensed phases. LSER with those theoretical parameters for log Poct, log Pc16 and log Pcic are chemically consistent. Atomic partial charges, widelly used to describe compounds and functional group reactivity in a qualitative way, can be used as a quantitative descriptor of molecular interactions in solubility, the first step wich drive a lot of chemical reactions.

Nesta tese investigamos computacionalmente clusters de Aln(+,0,-), n = 2-30, com especial interesse em entender aspectos fundamentais de suas propriedades em função do número de átomos e da carga total. Clusters de Aln(+,0,-) podem variar suas propriedades completamente com a adição de um átomo ou um elétron. Essas características permitem que atuem em nanocatálise, como ácidos ou bases de Lewis ou como substrato para adsorção de compostos orgânicos e inorgânicos. Apesar das propriedades dos clusters os tornarem ótimos materiais para inúmeras aplicações, a síntese desses materiais é desafiadora, podendo ser realizada por métodos químicos ou físicos. Os métodos físicos se baseiam na vaporização a laser, e em geral é utilizado um espectrômetro de massa para segregar as partículas, enquanto os métodos químicos compreendem a síntese por meio de nucleação e banhos químicos usando agentes redutores gerando clusters passivados ou suportados por substratos. Apesar de haver muitos estudos experimentais e computacionais investigando propriedades estruturais, eletrônicas e energéticas de clusters de Alumínio como distorção janh-Teller, energia de ionização e afinidade eletrônica, poucos foram realizados priorizando a análise da estrutura eletrônica visando investigar a relação entre suas propriedades e as ligações químicas. Dessa forma, o objetivo desse trabalho foi investigar as propriedades estruturais, energéticas e eletrônicas de clusters Aln(+,0,-) , n = 2 - 30, reunindo evidências que explicam a natureza de suas ligações químicas em função do número de átomos e da carga total. As propriedades estruturais compreendem distância média, número de coordenação efetivo e distorções Janh-Teller. As propriedades energéticas envolvem energia de estabilização por átomos, energia de dissociação e função de estabilidade. As propriedades eletrônicas englobam a energia de ionização, a afinidade eletrônica, os orbitais de fronteira, a densidade de estados, além da análise da densidade eletrônica por meio da teoria de átomos em moléculas (AIM), a análise decomposição energética por interação quântica entre átomos (IQA), a análise da função de localização eletrônica (ELF), a investigação da ordem de ligação pelo índice de Mayer, a análise da ligação química pelo método de particionamento da densidade natural adaptada (AdNDP) e análise dos orbitais naturais de ligações (NBO). As estruturas iniciais dos clusters de Aln(+,0,-) foram obtidas pelo algoritmo de busca configuracional Revised Basin-Hopping Monte Carlo (RBHMC), e posteriormente otimizados pelo modelo computacional TPSSh/def2-TZVPP, esta estratégia gerou resultados de energia de ionização e afinidade eletrônica mais próximos de valores experimentais que os modelos MP2/def2-TZVPP, PBE0/def2-TZVPP, BP/def2-TZVPP e B3LYP/def2-TZVPP. Nós observamos considerável influência da carga total e do número de átomos sobre a geometria dos clusters investigados. A distância de ligação média e o número de coordenação efetivo permitiram caracterizar clusters com estrutura planar, estrutura tridimensional, e também regiões de transição entre essas geometrias por meio da simples análise do gráfico dessas propriedades em função do número de átomos. A energia de ligação por átomos revelou que a estabilização desses clusters cresce rapidamente até Al12(+,0,-) , e a partir de então, se aproxima lentamente do valor da fase bulk conforme cresce o número de átomos. A função de estabilidade revelou que a estabilização de clusters está relacionada com a sua estrutura eletrônica, revelando que alguns clusters Aln(+,0,-), n = 2 - 30, são mais estáveis que os vizinhos mais próximos, em acordo com a sugestão dos números mágicos inicialmente aplicados para clusters de sódio e potássio. A energia de ionização e a afinidade eletrônica calculadas por TPSSh/def2-TZVPP para os clusters investigados forneceram resultados muito próximo dos valores experimentais, o comportamento oscilante dessas propriedades foram associados a padrões de crescimento baseado em icosaedros, decaedros e tetraedros por alguns autores e concordam com os nossos resultados. Os orbitais de fronteira e a densidade de estados dos clusters Al13(+,0,-) indicam que o HOMO apresenta contribuição preponderante dos orbitais pi. A teoria quântica de átomos em moléculas revelou que a força das interações químicas está associada à deslocalização eletrônica nos clusters investigados que cresce na seguinte ordem Aln+ < Aln0 < Aln-. E o caráter covalente nesses clusters é maior para os cátions que para os neutros e ânions. Esses resultados conferem maior caráter metálico aos clusters Aln- que aos demais clusters carregados e neutros, confirmados pelo método ELF. O método de partição energética IQA mostrou que a contribuição energética preponderante para a estabilização das interações químicas nos clusters Aln(+,0,-), n = 2 - 7, é a energia de troca responsável pelo caráter covalente, enquanto a contribuição iônica é oriunda da energia cinética eletrônica que apresenta caráter desestabilizante ou repulsivo para os clusters Aln(+,0,-), (n = 2 - 4 e 6 - 7), por outro lado, essa contribuição é estabilizante, mesmo que muito pouco para os clusters Al5(+,0,-) e consideravelmente estabilizante para as interações nos clusters Al13(+,0,-), sendo importante para o aumento da deslocalização eletrônica confirmada pela presença de aromaticidade tridimensional investigada pelos métodos ELF, AdNDP e NICS. / In this thesis we investigate computationally Aln(+,0,-) , n = 2 - 30 clusters with special interest in understand the fundamental aspects of theirs properties in function of number of atoms and of total charge. Aln(+,0,-) clusters can vary its properties completely through the addition of an atom or an electron. These features allow them to act at nanocatalysis such as acids or basis of Lewis or such as substrate for adsorption of organic and inorganic compounds. Despite the properties of clusters become them optimal materials for numerous applications, the synthesis of these materials is challenging, it can be performed by chemical or physical methods. Physical methods are based on laser vaporization and generally used a mass spectrometer to segregate the particles, while chemical methods include comprise through nucleation and chemical baths using reducing agents generating passivated clusters or supported by substrates. Although there are many experimental and computational studies investigating structural, electronic and energetic properties of Aluminum clusters such as Jahn-Teller distortion, ionization energy and electron affinity, few were performed prioritizing the analysis of the electronic structure and to assess the relationship between its properties and the chemical bonds. This form the objective this work was investigate the structural, energetic and electronic properties of Aln(+,0,-), n = 2 - 30 clusters gathering evidence to explain the nature of their chemical bonds in function of the number of atoms and of total charge. The structural properties include average distance, effective coordination number and Janh-Teller distortions. The energetic properties involve stabilization energy of atoms, dissociation energy and stability function. The electronic properties include the ionization energy and electron affinity, frontier orbitals, density of states, as well as analysis of the electron density through the atoms in molecules (AIM) theory, the analysis by interacting quantum atoms (IQA), the analysis of electron localization function (ELF), the investigation of binding order by Mayer index, the analysis of chemical bonding by adaptive natural density partitioning method (AdNDP) and analysis of natural bond orbital (NBO). The initial structures of clusters Aln(+,0,-) were obtained by algorithm of configuracional search Revised Basin-Hopping Monte Carlo (RBHMC), and subsequently optimized by computational model TPSSh/def2-TZVPP, this strategy performed electron affinity most close of experimental values than MP2/def2-TZVPP, PBE0/def2-TZVPP, BP/def2-TZVPP e B3LYP/def2-TZVPP. We observed considerable influence of the total charge and the number of atoms on geometry of the investigated clusters. The average bond distance and effective coordination number allowed the characterization of clusters such as planar structure, three-dimensional structure, as well as transition regions between these geometries through simple graphic analysis of these properties according to the number of atoms. The bind energy by atoms revealed that the stabilization of these clusters grows rapidly until Al12(+,0,-), and from them it grows slowly approaching from bulk phase, it grows in function of number of atoms. Stability function revealed that stabilization of clusters is related to its electronic structure indicating that some clusters Aln(+,0,-), n = 2 - 30 are more stable than the nearest neighbors, in accordance with the suggestion of the magic numbers initially applied to sodium and potassium clusters. The ionization energy and the electron affinity calculated within TPSSh/def2-TZVPP for the investigated clusters provided very close results from experimental values, the oscillating behavior of these properties have been associated with growth patterns based on icosahedrons, decahedrons and tetrahedrons by some authors and agree with our results. The frontier orbitals and the density of states of the Al13(+,0,-) clusters indicate that the HOMO has dominant contribution from orbitals. Quantum theory of atoms in molecules revealed that strength of chemical interactions is associated with the electron delocalization in the investigated clusters growing in the following order Aln+ < Aln0 < Aln-. The covalent character of these clusters is higher for cations than for neutral and anions. These results provide greater metallic character to Aln- clusters than others charged and neutral clusters, such as confirmed by ELF method. IQA energy partition method revealed that dominant energetic contribution to the stabilization of chemical interactions in clusters Aln(+,0,-), n = 2 - 7 is the exchange energy responsible by covalent character, while the ionic contribution comes from the electronic kinetic energy that has destabilizing or repulsive character to the Aln(+,0,-), (n = 2 - 4 and 6 - 7) clusters. On the other hand, this contribution is stabilized to Al5(+,0,-) clusters, even though very little, and considerable stabilizing to the interactions at Al13(+,0,-) clusters. Thus, it is important to increase of the electron delocalization confirmed by presence of the three-dimensional aromaticity investigated by ELF, AdNDP and NICS methods.

Les glaces du milieu interstellaire sont invoquées comme l'une des origines de la formation de molécules organiques complexes dans l'espace. En effet, elles constituent un support catalytique pour des réactions chimiques et pourraient ainsi expliquer la formation de molécules d'intérêt prébiotique. Toutefois, en raison de la faible température des milieux considérés, la vitesse de la réaction est contrainte par le déplacement des différentes réactifs l'un vers l'autre. L'objectif de cette thèse est donc de traiter la réactivité et la diffusion de molécules simples dans les glaces interstellaires. Je présente dans cette thèse les résultats de l'étude de la réaction entre le dioxyde de carbone et l'ammoniac dans les glaces interstellaires. Cette étude a été effectuée au Laboratoire Univers et Particules de Montpellier ; elle combine différentes méthodes de la chimie théorique et confronte les résultats avec ceux issus d'expériences que j'ai réalisées au Laboratoire de Physique des interactions Ioniques et Moléculaires de l'Université d'Aix-Marseille. Dans une première partie, des simulations de dynamique moléculaire classiques sont employées pour simuler un modèle de glace amorphe basse densité. Ce modèle est utilisé pour simuler la trajectoire de petites molécules (NH3, CO, CO2, H2CO) dans cette glace et en déduire des coefficients de diffusion à différentes températures. Ces résultats sont comparés à des résultats expérimentaux de diffusion du dioxyde de carbone ce qui valide la méthode théorique utilisée et permet de suggérer un mécanisme pour ce processus de diffusion. Dans une deuxième partie, la réaction entre le dioxyde de carbone et l'ammoniac est traitée dans le cadre de la théorie de la fonctionnelle densité par une approche « super-molécule ». Dans cette approche, le profil d'énergie et le mécanisme de la réaction dans des complexes moléculaires xNH3:CO2:yH2O sont étudiés. Deux produits de la réaction sont localisés : le carbamate d'ammonium et l'acide carbamique. La barrière d'énergie de la réaction obtenue est similaire à celle obtenue expérimentalement, et le carbamate d'ammonium est confirmé comme produit majoritaire de la réaction. Le profil d'énergie obtenu par cette approche « super-molécule » est ensuite étudié par dynamique moléculaire ab initio contrainte et le profil d'énergie libre est calculé par la méthode d'Intégration Thermodynamique. Cette approche confirme la forme générale du profil d'énergie et met en évidence un fort effet entropique du réseau d'eau. Je dresse finalement des conclusions sur les plans méthodologiques et astrochimiques permettant d'inscrire ma thèse dans des perspectives futures, notamment en incluant les barrières d'énergie de diffusion et de réaction, dans des modèles astrochimiques prenant en compte directement les réactions chimiques dans les glaces interstellaires. / It is postulated nowadays that complex organic molecules in space form on the surface and in the volume of interstellar ices. These ices can catalyse chemical reactions what could explain the formation of prebiotic molecules. However, because of the low temperatures, the diffusion of the reactants one towards another is slow, limiting their reactivity. The objectif of this thesis is to treat the reactivity and the diffusion of simple molecules in interstellar ices. I present in this thesis the results of the study of the chemical reaction between carbon dioxide and ammonia in interstellar ices. This study was conducted in the Laboratoire Univers et Particules de Montpellier : it combines various theoretical chemistry methods and confront the results to experimental ones I participated in at the Laboratoire de Physique des Interactions Ioniques etMoléculaires of the Aix-Marseille university. In a first part, classical molecular dynamic simulations are used to simulate a low-density amorphous ice model and to calculate the diffusion coefficients at various temperatures of a series of small molecules (NH3, CO, CO2, H2CO). These results are compared to the experimental diffusion coeficients of the carbon dioxide validating the theoretical approach used and allowing to porpose a mechanism for the diffusion process. In a second part, the reaction between carbon dioxide and ammonia is studied in the frame of Density-Functional Theory using a « super-molecular » approach. In this approach, the reaction energy profile for the molecular complexes xNH3:CO2:yH2O, is studied. Two reaction products are localised : the ammonium carbamate and the carbamic acid. The reaction energy barrier obtained by these calculations is similar to the one obtained experimentally, and the ammonium carbamate is confirmed as the major reaction product. The obtained energy profile is therfore investigated by constrained ab initio molecular dynamics and the free energy profile is computed with the Thermodynamics Integration method. These calculations confirme the general form of the previous energy profile and enlight the strong entropic effect of the water network. Finally, conclusions are drawn, and perspectives on methodological as well as on astrochemical aspects, as the inclusion of the reaction and diffusion energy barriers I calculated in astrochemical models, are given.

As another important features in VEELS, the plasmon excitations (including the volume plasmon and surface/interfacial plasmon) are also utilized to identify different phases and multi compositions within materials. The microstructure and electronic structure evolution of silicon-rich oxide (SRO) films as a function of the annealing temperature are investigated using TEM and VEELS. The as-deposited SiO film is found to be a single phase with only single volume plasmon presents in VEEL spectrum and almost no interfacial plasmon is observed. After the annealing (Tanneal>400&deg;C), it begins to decomposite into Si and SiO2 and the single phase changed into cluster/matrix nanocomposites where the interfacial plasmon appears. The Si duster size and its concentration increase as the annealing temperature increases. / Firstly, the applications of VEELS in investigating the electronic structures of ZnO nanowires with different diameter and surface shapes are demonstrated. Using the momentum transferred technique, one of the interband transitions with dipole-forbidden nature is identified. Several size dependent features are found on the interband transitions and plasmon oscillations of ZnO nanowires with small diameter and circular cross section, which are mainly due to the large surface to volume ratio and existence of Oxygen dangling bonds on those ZnO wires. / Further explorations on the electronic structure in the vicinity of band gap are carried out for the ZnO nanowires doped with different dopants (Co, Er, Yb) and different dopant concentrations. In order to obtain trustworthy information in the very low energy range of VEELS a narrow zero loss peak and elimination of Cerenkov effect and surface losses are necessary, which can be realized by incorporation of the gun monochromator in the TEM and taking spectrum at a momentum transfer slightly greater than zero. Band tail states (&sim;2-3.3 eV) are found to be generated in the ZnO nanowires after the ion implantation and their density of states increase with the ion fluence increases. The partially removal of those defect states by the Oxygen annealing is also observed in VEELS. On the other hand, interesting mid-gap state(s), which is dopant-sensitive (as it is only observed in the rare earth (Er and Yb) doped ZnO nanowires, but not in the Co-doped ones), does not show obvious change after the O annealing. The impact of these electronic structure changes on the material properties are also discussed. / In the end of the thesis, some of the practical limitations and contradictories on the energy resolution (DeltaE), spatial resolution (Delta x), and the momentum resolution (Deltaq) when carrying out the various VEELS study are summarized. The compromise made among these resolution limits is also discussed. / In this work, the important experimental parameters and appropriate data processing methods to generate trustworthy data are discussed. Based on that, three material systems, i.e., pure ZnO nanowires, doped ZnO nanowires, and Si/SiO/SiO2 composite films are investigated. Various information on the material microstructure/electronic structure is interpreted using the VEELS data. / The valence-electron energy-loss spectroscopy (VEELS) contains information on the electronic structures of materials, including the band gap the single-electron interband transitions and the plasmon oscillations. When operating in transmission electron microscope (TEM), the excellent spatial resolution enables the VEELS not only exploring the local electronic structures of individual low dimensional nanostructured materials, but also building up correlations between the electronic structure and microstructure. In addition, the capability in carrying out the momentum transfer dependent study in VEELS allows the investigation on the dispersion of plasmons and single electron excitations in the momentum space. The optically forbidden transitions, which are not allowed in conventional optical method, can also be excited at high momentum transfer values using VEELS. / Wang, Juan = 價電子能量損失譜在低維納米材料中的應用 / 王娟. / "September 2007." / Adviser: Li Quan. / Source: Dissertation Abstracts International, Volume: 69-02, Section: B, page: 1267. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2007. / Includes bibliographical references (p. 122-133). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract in English and Chinese. / School code: 1307. / Wang, Juan = Jia dian zi neng liang sun shi pu zai di wei na mi cai liao zhong de ying yong / Wang Juan.

Parâmetros teóricos obtidos pela otimização de geometria da estrutura molecular em fase gasosa com o método quântico semi-empírico PM3, foram usados para correlacionar constantes de partição (log Poct; log Pc16 e log Pcic) para uma série de 700 solutos, considerando como referência as LSER obtidas com os parâmetros experimentais dos solutos: acidez (&#931;&#945;2), basicidade (&#931;&#946;2), e dipolaridade/polarizabilidade (&#960;2) de ligação de hidrogênio; refração molar de excesso (R2) e volume característico (Vx). Os parâmetros avaliados foram: as energias HOMO e LUMO; o momento de dipolo elétrico e as cargas atômicas parciais do hidrogênio mais positivo e do átomo mais negativo (classe IV - modelo CM1 ); e o índice de polarizabilidade como proposto por Famini e col.. Estes parâmetros teóricos, juntamente com Vx, são adequadas para serem utilizados em LSER na correlação dos valores de log P. Para a maioria dos solutos considerados, as quantidades eletrostáticas q+ , &#921;q-&#921; e &#181; descrevem adequadamente os termos de acidez, basicidade e dipolaridade de ligação de hidrogênio, o que não ocorre para solutos com mais de um polo positivo ou negativo relevante, como hidroquinona, nem para solutos como 2-nitrofenol com capacidade de formar ligação de hidrogênio intramolecular. As LSER para log Poct; log Pc16 e log Pcic com estes parâmetros teóricos são quimicamente consistentes. Cargas atômicas parciais, largamente usadas qualitativamente para descrever a reatividade de compostos e de grupos funcionais, podem ser usadas como descritor quantitativo de interações moleculares em estudos de solubilização, o primeiro estágio, que governa muitas reações químicas. / Semiempirical quantum theoretical parameters, obtained from PM3 geometry optimization method in gas-phase for a wide range of 700 solutes, were used to correlate experimental partition coeficients (log Poct; log Pc16 and log Pcyc), bearing as reference the LSER with experimentally derived parameters: Hydrogen bond acidity (&#931;&#945;2), basicity (&#931;&#946;2) , and dipolarity/polarizability (&#960;2); excess molar refraction and characteristic volume (Vx). Theoretical quantities tested includes: HOMO and LUMO energies; class IV, CM1 model atomic partial charges (q+ and &#921;q-&#921;) and dipole moment (&#181;); Famini\'s polarizability index (&#960;F); and Vx. Those quantum derived quantities can successfully be used, jointly with Vx, to correlate log P values in the LSER approach. For most of the considered solutes, electrostatic quantities q+, &#921;q-&#921; and &#181; describes fairly well Hydrogen bond acidity, basicity and dipolarity terms, but do not describe hydrogen bond terms for solutes with more than one relevant positive or negative centers, like hydroquinone, nor solutes like 2-nitrophenol wich forms intramolecular hydrogen bond when in condensed phases. LSER with those theoretical parameters for log Poct, log Pc16 and log Pcic are chemically consistent. Atomic partial charges, widelly used to describe compounds and functional group reactivity in a qualitative way, can be used as a quantitative descriptor of molecular interactions in solubility, the first step wich drive a lot of chemical reactions.