The theory of non-orthogonal spin-adaptation for closed-shell molecular systems is presented, with an emphasis on application to the coupled cluster family of electronic structure methods. To aid in the derivation of efficient and compact working equations, a new diagrammatic interpretation of the Goldstone diagrams is derived which only requires a small number of the many distinct diagrams and which directly produces equations in a factored form in terms of “spin-summed” tensor elements. This diagrammatic interpretation is applied to coupled cluster methods with quadruple excitations (CCSDTQ), including coupled cluster with a perturbative correction for quadruple excitations (CCSDT(Q)) and to CCSDTQ gradients and properties. The advantages of the non-orthogonal spin-adaption with respect to simplification and factorization of the working equations and to efficient implementation are presented and discussed. Additionally, specific optimizations of the implementation for often-overlooked issues such as tensor transposition, disk access, and removal of redundant and/or unnecessary operations are detailed. The resulting algorithm is implemented for the CCSDTQ and CCSDT(Q) methods and compared to existing codes, where a one to two order-of-magnitude improvement in efficiency is observed. The new implementation is also used for calculations on several larger molecular systems to illustrate the scalability of the method. / text

No description available.

The electronic structure factors that control Si-H and Ge-H bond activation by transition metals are investigated by means of photoelectron spectroscopy. Molecular orbital calculations are also used to gain additional insight into the orbital interactions involved in bond activation. The complexes studied have the general molecular formula (η⁵-C₅R'₅)Mn(CO)(L)HER₃, where R' is H or CH₃, L is CO or PMe₃, E is Si or Ge and R is Ph or Cl. These compounds are interesting models for catalysts in industrial processes like hydrosilation. The compounds display different stages of interaction and "activation" of the E-H bonds with the metal. One purpose is to measure the degree of Mn, Si, H 3-center-2-electron bonding in these complexes. The three-center interaction can be tuned by changing the substituents on Si, methylating the cyclopentadienyl ring, changing the ligand environment around the metal and substituting Si with Ge. The degree of activation is measured by observing the shifts in the metal and ligand ionizations relative to starting materials and free ligand in the photoelectron spectrum. Changing the substituent on Si extensively changes the degree of activation. Photoelectron spectral studies on (η⁵-C₅H₅)Mn(CO)₂HSiPh₃ show this to be a Mn(I) system. Progressive methylation of the cyclopentadienyl ring increases the electron richness at the metal center with no substantial effect on the degree of activation. Substitution on the metal (PMe₃ for CO) is less able to control the electronic structure factors of activation than the substitution on the Si atom. The magnitude of Ge-H bond activation is found to be of the same order as the Si-H bond activation for analogous compounds as found by studying (η⁵-C₅H₅)Mn(CO)₂HGePh₃, (η⁵-CH₃C₅H₄)Mn(CO)₂HGePh₃ and (η⁵- C₅(CH₃)₅)Mn(CO)₂HGePh₃ complexes by photoelectron spectroscopy. The photoelectron spectra of CpFe(CO)₂SiCl₃ and CpFe(CO)₂SiMe₃ were measured to study the electron charge shift from the metal to the ligand in these complexes as compared to CpMn(CO)₂HSiR₃ complexes. The photoelectron spectroscopic studies include numerous perturbations of the ligand and metal center to observe the extent of bond interaction and remain one of the best techniques to detect activation products.

This dissertation describes the results of time-of-flight spectroscopic examination of H⁺ ions resulting from electron bombardment of the hydrogen halide molecules HF, HCl, HBr, and HI. The time-of-flight spectra of the H⁺ fragments and their corresponding H⁺ fragment kinetic energy spectra are used to study the dissociative processes that yield H⁺ fragments for electron bombardment energies in the 15 eV to 51 eV range. The H⁺ fragments are produced in an interaction region defined by a pulsed electron beam colliding with the target gas. By keeping the gas pressure sufficiently low to guarantee that the fragment path length to the ion detector is much less than the mean-free path length in the gas, the fragments' velocities can be considered a sample of fragment velocities produced by the electron beam and hydrogen halide gas in the interaction region. The geometry of the interaction region primarily detected fragments produced at 90° to the electron beam axis. The electron gun used was designed to allow computer control of the electron bombardment energy. The computer also controlled a programmable multichannel analyzer that allowed the data to be acquired in a fashion that permitted normalization of the H⁺ TOF spectra taken at different electron bombardment energies. This normalization procedure allowed the use of ionization efficiency curves in detection of the thresholds of H⁺ production channels for HCl and HBr. For HF and HI the thresholds of H⁺ production channels had to be determined by visual examination of the TOF spectra. The electronic structure of the hydrogen halide molecules has been a popular topic of study over the years. Since this work represents the first TOF study of electronic excitation processes that lead to dissociation resulting in H⁺ fragments from the hydrogen halides, it should prove to be a significant contribution toward an understanding of the highly excited electronic states of these molecules and their molecular ions. The interpretation of the results obtained indicated that both configuration interactions between adiabatic electronic states that lead to predissociation-type processes and inner valence shell excitations were probably the primary contributors to the H⁺ fragment production.

No description available.

No description available.

No description available.

A thorough understanding of the underlying electronic principles guiding the stability and reactivity of clusters has direct implications for the identification of stable clusters for incorporation into clusters-assembled materials with tunable properties. This work explores the electronic principles governing the stability and reactivity of two types of clusters: ligated metal clusters and silicon encapsulated transition metal clusters. In the first case, the reactivity of iodine-protected aluminum clusters, Al13Ix- (x=0-4) and Al14Iy- (y-0-5), with the protic species methanol was studied. The symmetrical ground states of Al13Ix- showed no reactivity with methanol but reactivity was achieved in a higher energy isomer of Al13I2- with iodines on adjacent aluminum atoms – complementary Lewis acid-base active sites were induced on the opposite side of the cluster capable of breaking the O-H bond in methanol. Al14Iy- (y=2-5) react with methanol, but only at the ligated adatom site. Reaction of methanol with Al14- and Al14I- showed that ligation of the adatom was necessary for the reaction to occur there – revealing the concept of a ligand-activated adatom. In the second case, the study focused heavily on CrSi12, a silicon encapsulated transition metal cluster whose stability and the reason for that stability has been debated heavily in the literature. Calculations of the energetic properties of CrSin (n=6-16) revealed both CrSi12 and CrSi14 to have enhanced stability relative to other clusters; however CrSi12 lacks all the traditional markers of a magic cluster. Molecular orbital analysis of each of these clusters showed the CNFEG model to be inadequate in describing their stability. Because the 3dz2 orbital of Cr is unfilled in CrSi12, this cluster has only 16 effective valence electrons, meaning that the 18-electron rule is not applicable. The moderate stability of CrSi12 can be accounted for by the crystal-field splitting of the 3d orbitals, which pushes the 3dz2 orbital up in energy. CrSi14, on the other hand, has 18 effective valence electrons on Cr, minimal 3d-orbital splitting, and does follow the 18-electron rule. A repetition of these calculations with WSin (n=6-16) showed similar results, except WSi12 shows all the markers of a magic cluster, due to the greater crystal-field splitting of 5d orbitals.

L’étude de réactions chimiques, tout comme le calcul de propriétés thermodynamiques, sont des enjeux capitaux de la chimie moderne. L’évolution des instruments et techniques expérimentales permet des mesures de plus en plus précises de ces grandeurs, pour des systèmes de plus en plus complexes. L’intérêt croissant pour l’étude du milieu interstellaire et des atmosphères planétaires se révèle également être un défi très important dans les décennies à venir. Les difficultés rencontrées lors de l’analyse de ces expériences (ou mesures), nécessitent souvent l’intervention de simulations numériques de manière à éclairer ces observations. Une autre utilisation du calcul est de prédire des paramètres moléculaires et spectroscopiques d’espèces instables difficiles à produire au laboratoire. Les outils actuels de la chimie théorique ab initio sont des moyens précieux pour la prédiction et l’interprétation de résultats expérimentaux ou de mesures astrophysiques et atmosphériques. Ces techniques de simulation ont connu des développements importants au cours des dernières décennies. Les progrès récents en matière de calculs d’interaction de configurations de grande taille permettent d’inclure une grande partie de l’énergie de corrélation. Le temps de calcul et la taille mémoire des ordinateurs restent cependant des limites importantes qui ne permettent pas d’effectuer des interactions de configurations totales dans une base suffisamment grande pour contenir la physique des systèmes étudiés au delà de petites molécules. Cet état de fait conduit à s’intéresser à des méthodes moins coûteuses comme celles des perturbations, les interactions de configurations tronquées et le Coupled Cluster, permettant d’inclure une partie de la corrélation électronique à un coût moins élevé en temps de calcul. Ce sont ces méthodes qui ont été utilisées dans ce travail pour déterminer théoriquement les paramètres moléculaires et spectroscopiques des systèmes MgO, MgO+, FeC2, FeC2+ et FeC2- avec le maximum de précision possible.Dans un premier temps, nous avons étudié la molécule MgO. C’est un système de choix car, il permet de s’initier aux méthodes de calcul ab initio sur les systèmes moléculaires les plus simples (diatomiques), de tester et de comprendre ces méthodes (différentes approximations, validité, précision, …) et de bien interpréter les résultats obtenus (formation de la liaison chimique et des états moléculaires, leur symétrie, leurs couplages, leur stabilité, leur spectroscopie, …) surtout qu’il a fait l’objet de plusieurs études théoriques et expérimentales. Pour profiter de notre savoir-faire pour les molécules diatomiques nous avons étudié le système MgO+ qui a fait l’objet de notre deuxième article que sera présenté en annexe.Dans un second temps, nous avons visé les systèmes moléculaires de type FenCm afin de comprendre la croissance et la dynamique des nanotubes de carbone catalysée par le Fer. Le système diatomique FeC fait l’objet de plusieurs études théoriques et expérimentales. La plus récente est celle fourni par Demeter Tzeli et Aristides Mavridi. Cette étude théorique a caractérisé son état fondamental ainsi que les 40 états électroniques les plus bas, à toutes les distances internucléaires jusqu’à la dissociation, et d’autre part de fournir des données spectroscopiques d’une précision comparable à celle donnée par l’expérience. Pour les systèmes d’ordre supérieur, confronté par le problème que ces petits systèmes moléculaires constitués de Fer et de Carbone ont des structures électroniques très compliquées, notre étude s’est limitée à l’étude des systèmes FeC2, FeC2+ et FeC2- / The study of chemical reactions, as well as the calculation of thermodynamic properties are critical issues of modern chemistry. The development of experimental techniques and instruments allows measurements more accurate these quantities for systems more complex. The growing interest in the study of the interstellar medium and planetary atmospheres is also proving to be a major challenge in the coming decades. The difficulties encountered in the analysis of these experiences (or measures) often require the intervention of numerical simulations to clarify these observations. Another use of the calculation is to predict molecular and spectroscopic parameters of unstable species are difficult to produce in the laboratory.Current tools of theoretical chemistry ab initio are valuable tools for the prediction and interpretation of experimental results or astrophysical measurements and atmospheric. These simulation techniques have experienced significant developments in recent decades. The recent progress in calculations of interaction of large configurations can include a large part of the correlation energy. The computation time and memory size of computers, however, remain significant limitations that do not allow to perform configuration interaction in a total base large enough to hold the physical systems studied beyond small molecules. This fact led to interest in cheaper methods such as disruption, the truncated configuration interaction and coupled cluster, allowing to include a portion of electron correlation at a lower cost in computation time. These are methods that have been used in this work to determine theoretically the molecular parameters and spectroscopic systems MgO, MgO +, FEC2, FEC2 + and FEC2-with maximum accuracy.As a first step, we studied the MgO molecule. It is a system of choice because it allows you to learn the methods of ab initio calculations on molecular systems the simplest (diatomic), test and understand these methods (different approximations, validity, accuracy, ...) and to properly interpret the results (formation of chemical bonding and molecular states, their symmetry, their interactions, their stability, spectroscopy, ...) especially since it has been the subject of several theoretical and experimental studies. To take advantage of our expertise for diatomic molecules we have studied the system MgO + has been our second article will be presented in the appendix.In a second step, we targeted molecular systems FenCm like to understand the growth and dynamics of carbon nanotubes catalyzed by iron. Diatomic system FeC been several theoretical and experimental studies. The most recent is provided by Demeter Tzeli and Aristides Mavridi. This theoretical study has characterized its ground state and the 40 lowest electronic states at all internuclear distances up to dissociation, and secondly to provide spectroscopic data with an accuracy comparable to that given by the experiment. For higher-order systems, the problem faced by these small molecular systems composed of iron and carbon have very complicated electronic structures, our study is limited to the study of systems FEC2, FEC2 + and FEC2-

Neste trabalho apresentamos um procedimento para o cálculo da contribuição eletrônica para o gradiente de campo elétrico no núcleo (GCE), baseado no método de recorrência e no método de arbitais \"muffin-tin\" linearizados (LMTO) na representação forteme~ te ligada (\"tlght-binding\"). Este esquema é desenvolvido no esp~ ço real e pode ser aplicado em sistemas sem simetria, tais como metais amorfos, onde métodos de espaço-k não podem ser utilizados. A contribuição eletrônica é a dominante em sistemas metálicos e requer para sua avaliação o conhecimento das funções de onda eletrônicas do sistema. O esquema sugerido aqui, possui duas etapas principais: a es colha da Hamiltoniana do sistema, que neste caso foi a LMTO na aproximação de esfera atômica fortemente ligada (LMTO-ASA-TB), e o desenvolvimento dos termos do tensor GCE em uma abordagem de orbl tais fortemente ligados. Para o cálculo deste último, sugerimos a utilização do método de recorrência na obtenção dos elementos diagonais e não diagonais da função de Green, que compõem as expressões dos termos do tensor GGE em cada sitio. Como teste para nosso procedimento de espaço real, calculamos o GCE para o ao longo do eixo Zr hcp com o eixo /1. z do sistema de /1. c do cristal e também ao longo de coordenadas uma direção arbitrária. Os resultados se mostraram independentes da escolha do sistema, mostrando que o procedimento funciona na ausência de simetria. FinaJ.mente, como iLustração, calcuLamos a distribuição de GCE\'s para um agJ.omerado de Zt amorfo construído segundo o mode- Io de empacotamento denso e aleatório de esferas rígidas (DRPHS). / We present here a new method to cal-culate the el-ectronic contrlbution to the electric field gradient (GCE), which is based on the recursion method and in the recently developed tight-binding representation of the LMI0 method. This is a real-space scheme and can be applied to systems which lack symmetry, as amorphous metal.s where R-space methods can not be used. The electronic contribution is dominant Ín metalLÍc systems and requÍres for its evaluation the knowledge of the electronic wave functions of the system. The scheme I¡/e suggest here has two main steps: the choice of the Hamiltonian which in this case was the tight-binding LMT0 i\'n the atomÍc sþhere approxÍmation ( t-Uf O-ASA-TB ) and the deveJ.opment of the GCE tensor\'s components in a tight-binding approach. In the last step we suggest to use the recursion method to caLculate the diagonal and the off-diagonal Greenrs function eLements which build the GCE tensot\'s components in each site. As a test of our real space method we calculate the GCE for hcp zr with the c axis parall-eL to the crystaI\' s z axis and also in an arbitrary direction. The resuLts have shown to be independent of the choice of the system, showing that the procedure works in the abcense ofl symmetry. FinaJ.ly, as an iliustration we cal-cuLate the GCErs distribution for an amorphous Zr cluster buÍ1t in the model of dense random packing of hard spheres (DRPHS).