ESI-MS spectra of Ni(II), Co(III), Mg(III), and Fe(II) porphyrin solutions in methanol show porphyrin monomer species with different charge states, such as [Ni(II)TPPS+H]3-, [Co(III)TPPS]3-, [Mn(III)TPPS]3-, [Mn(III)TPPS+H]2-, [Fe(II)TPPS+H]3-, and [Fe(II)TPPS+2H]2- ions. Collision-induced dissociation (CID) of these monomer species produced primarily losses of neutral SO3 and SO2. The mechanisms, in which these dissociation pathways took place, were investigated by the means of DFT calculations of the corresponding dissociation of neutral and ionized benzenesulfonate (B3-LYP/6-31+G(2d, p) level) and porphyrin monomer (B3-LYP/6-31+G(2d, p)+LANL2DZ//PM7 level). RRKM fitting of the CID breakdown curves showed that the activation energies of the reactions that experience a loss of SO2 from [Co(III)TPPS]3- and [Mn(III)TPPS]3- were similar, but of a lower magnitude than those for a loss of SO3. On the other hand, for [Ni(II)TPPS+H]3- and [Fe(II)TPPS+2H]2-, the activation energies of the reaction leading to a loss of SO2 were also similar, but this time were larger than those leading to SO3 loss. These results are consistent with a mechanism by which the SO2 loss starts with -C6H4SO3-, while the SO3 loss has to begin with -C6H4SO3H. To lose this SO3, extra energy is required for [Co(III)TPPS]3- and [Mn(III)TPPS]3- in order for them to overcome the barrier of H transfer from the porphyrin ring to -SO3-, but this is irrelevant when it comes to [Ni(II)TPPS+H]3- and [Fe(II)TPPS+2H]2- since the C6H4SO3H moiety already exists. In addition, the reaction of [Fe(II)TPPS+H]3- losing H leads to a unique dissociation mechanism.

In the last several decades, polycyclic aromatic hydrocarbons (PAHs) have been the subject of extensive investigation due to their presumed abundance in the interstellar medium (ISM). My thesis concentrated on investigating the dissociation of ionized PAH molecules in the gas phase under similar conditions of the ISM. The twelve PAHs studied were naphthalene (NAP), anthracene (ANT), phenanthrene (PHN), cyclopenta[d,e,f] phenanthrene (CYC), pyrene (PYR), fluoranthene (FLN), perylene (PER) and coronene (COR). In addition, two dihydro PAHs were studied to examine potential intermediates in H2 catalysis, 1,2-dihydronaphthalene (DHN) and 9,10-dihydrophenanthrene (DHP), and two stabilized fragments of PAHs, acenaphthylene (ACE), fluorene-H (FLU). These ions were studied using atmospheric pressure chemical ionization (APCI) to generate the ions and the fragmentation was produced by collision-induced dissociation (CID). The CID experiments were done at different lab frame collision energies to produce breakdown curves for all fragments in each molecule. These curves were fitted by using Rice – Ramsperger-Kassel- Marcus (RRKM) theory to derive the activation energy (E0) at 0K and the entropy of activation (∆‡ S) for each reaction. The primary dissociation channel observed was H loss and this fragmentation channel was a common fragment to all PAHs molecules studied. Also, there were other fragments (CH3, C2H2, C4H2) observed in all of these molecules studied except perylene and coronene. The final step in this project was to compare these results with results derived from imaging photoelectron photoion coincidence spectroscopy (iPEPICO) reported by West and co-workers in order to determine the suitability of the APCI-CID technique for deriving reliable kinetic parameters for this class of ions.

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-05-20T12:56:17Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Ayyaz-Mahmood.pdf: 9023674 bytes, checksum: beecc131e286e19ce6d03dec91f319d8 (MD5) / Made available in DSpace on 2016-05-20T12:56:17Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Ayyaz-Mahmood.pdf: 9023674 bytes, checksum: beecc131e286e19ce6d03dec91f319d8 (MD5) Previous issue date: 2015-11-27 / Realizamos estudos do mecanismo, da seletividade e regiosseletividade da metilação de nitronatos [R1R2CNO2]− com CH3I, em que R1 = R2 = H (1), R1 = CH3 e R2 = H (2), R1 = R2 = CH3 (3); e R1 + R2 = c-(CH2)2 (4), usando método quântico ab initio MP2/CBS método e as teorias ADO e RRKM e simulações numéricas. As estruturas de todos os pontos estacionário no perfil da reação, incluindo os estados de transição, foram obtidas com sucesso e forneceram as energias de ativação (Δ‡) e de reacção (Δr) para todos os caminhos reacionais. Estas reações iniciam com a formação de complexo de reagente do tipo íon-molécula (RC@C ou RC@O) que se converte no respectivo complexo de produto (PC@C e PC@O), através do estado de transição (TS@C e TS@O) correspondente. As constantes de velocidade para a conversão do RC ao PC foram calculadas com a teoria RRKM. Ao contrário da proposta da literatura para o mecanismo de reação, a C-metilação é a via termodinâmica e cinética preferencial, com as seguintes seletividades: PC@C:PC@O = 89:11 (1), 97:03 (2-E), 97:03 (2-Z), 100:0 (3) e 100:0 (4). Isto é corroborado pela concordância qualitativa e quantitativa entre a ordem da reatividade calculada e experimental: 4 >> 3 > 2 > 1. A regiosseletividade para a C-alquilação é explicada pela maior exotermicidade desse canal de reação em comparação com a O-alquilação. O aumento da reatividade com o aumento do impedimento estérico ao redor do carbono central é explicado pelas diferenças de carácter sp3 deste átomo e pela diminuição da frequência vibracional associada à piramidalização ao redor deste átomo de carbono. Uma avaliação de funcionais híbridos- e duplo-híbrido DFT e do método MP2 para estas reações foi realizada a fim de verificar a confiabilidade e precisão destes métodos e sua utilização em modelagens similares. Os métodos mais precisos e robustos, que apresentam menor demanda computacional, foram M06- 2X and MP2 com as funções de base aug-cc-pVTZ-PP. A fim de revelar os efeitos solvente e do agente alquilante, a metilação do nitronato [MeCHNO2]− (2) com Me3O+ (oxônio) foi estudada em fase gás e em solução com o método B2PLYP/6-311+G(d,p). Os resultados mostraram regiosseletividade para a O-metilação em fase gás e principalmente em solução (diclorometano), o que concorda com os dados experimentais. Além disso, eles mostram que a regiosseletividade é fortemente dependente do agente alquilante (MeI vs. Me3O+). Este mesmo método foi utilizado para as reações [XC6H4CHNO2]− + MeCl (X = H, p-MeO, p-NO2) em fase gás e em solução (DMF). Os resultados em solução mostram uma pequena seletividade favorável à O-metilação, contudo, considerando que o agente metilante usado experimentalmente era o metil-p-bromobenzenossulfonato (MeOBS), o qual é muito mais reativo que o MeCl, é esperado que a utilização do MeOBS nas modelagens moleculares deva aumentar esta regiosseletividade para a O-metilação. / We performed a mechanistic, selectivity and regioselectivity study of methylation of nitronates [R1R2CNO2]− by CH3I in the gas-phase, where R1 = R2 = H (1), R1 = CH3 and R2 = H (2), R1 = R2 = CH3 (3); and R1 + R2 = c-(CH2)2 (4), using ab initio MP2/CBS method, ADO and RRKM theories and kinetics simulations. The structures of all stationary points, including transition states, were successfully obtained to provide energies of activation (Δ‡) and of reaction (Δr) for all reaction pathways. These reactions proceed with the formation of ion-neutral reactant complex (RC@C or RC@O) that is converted into the respective product complex (PC@C and PC@O) through the corresponding transition state (TS@C and TS@O). The rate constants for the conversion of RCs into PCs were calculated with the RRKM theory. Contrary to a previous proposal for the reaction mechanism, the C-methylation is the thermodynamics and kinetics preferred pathway with the following PC@C:PC@O selectivites 89:11 (1), 97:03 (2-E), 97:03 (2- Z), 100:0 (3), and 100:0 (4). This is corroborated by the agreement between the calculated and experimental reactivity trend 4 >> 3 > 2 > 1. The regioselectivity towards the C-alkylation is explained by the much larger exothermicity of this reaction channel compared to the Oalkylation. The increase of reactivity upon the increase of crowdedness of the central carbon atom is explained by differences of sp3 character at this atom and the decrease of the vibrational frequency associated with pyramidalization around this carbon atom. An assessment of hybrid- DFT functionals and MP2 method for these reactions was performed in order to ascertain the reliability of these methods for similar molecular modeling. The most accurate and robust methods, with the least computational demand, were M06-2X and MP2 with aug-cc-pVTZ-PP basis sets. In order to reveal the solvent and alkylating agent effects, the methylation reaction [MeCHNO2]− (2) + Me3O+ (oxonium) was studied in the gas-phase and in solution with B2PLYP/6-311+G(d,p) calculations. The results show regioselectivity towards O-methylation in the gas-phase and more significantly in solution (dichloromethane), which is in agreement with the experimental observation. In addition, it shows that the regioselectivity is highly dependent upon the reactivity of the alkylating agent (MeI vs. Me3O+). This same method was employed for the [XC6H4CHNO2]− + MeCl (X = H, p-MeO, p-NO2) reactions in the gas-phase and in solution (DMF). The results in solution show a slight regioselectivity towards O-methylation; however, considering that the methylating reagent used in the experiments was methyl pbromobenzenesulfonate (MeOBS), which is much more reactive than MeCl, it should thus be expected an increase of the calculated selectivity.

In the technologically driven information age, the consumption of power is as vital to daily life as food and shelter. The generation of that power comes from a variety of sources of which coal is the predominant provider of electrical energy. Coal combustion is a well-known technology and the United States possesses the most abundant coal deposits on Earth, however, the drawback accompanying this process is the significant emissions which are released during combustion. Over the years, much effort has gone into reducing the emissions of majority constituent elements CO2, CO, NOx, SOx, etc. but it is only in the last decade or so that much attention has been given to the trace metals present within coal. Most of the work into examining these trace metals has been upon Hg and how it speciates within the flue gas in order to determine the most effective means of removal. In this study, the trace metals arsenic (As) and selenium (Se) will be investigated in a similar manner to evaluate the speciation of these elements. While previous experimental work has been performed in this area, it has been limited to thermodynamic studies which determine the speciation after equilibrium has been reached, this ignores the fact the residence times within the flue are often only several minutes during which time rapid quenching is taking place. This study takes a different approach by examining the speciation using computational chemistry which affords the advantage of being able to perform a kinetic study which is more useful in creating a flue gas model. Using ab initio the properties of various As and Se species can be evaluated compared to existing experimental data for validation. After which, a number of reactions may be selected and the structure of the transition state for each identified. Once the properties of the transition structure are known, the appropriate kinetic model, be it Transition State Theory, RRKM Theory, etc. can be applied and the rate constant determined. It is by the determination of these rate constants that the kinetic model of the flue gas can be improved and a more accurate depiction of the speciation of these race metals created.

As a first project, collision-induced dissociation experiments were carried out using electrospray ionisation mass spectrometry on gas phase complexes involving different poly(methylmetacrylate) oligomers with three amino acids: glycine, leucine, and phenylalanine. After acquiring breakdown diagrams, RRKM modeling was used to fit the experimental data in order to obtain the 0 K activation energy and the entropy of activation. These thermodynamic data were then used to understand the competing dissociation channels observed (except for gas phase complexes involving glycine that had only one dissociation channel). Molecular dynamics simulated annealing calculations were carried on the gas phase complexes to understand further the energetic and entropic effects involved as well as the 3D conformation of these complexes. Valuable insight information was found on the 3D conformations, on a qualitative level. Using rotational constants and vibrational harmonic frequencies, it was possible to evaluate the entropy variation between the experimentally observed competing channels. Reasonable agreement was found between the experimental and theoretical variations of entropies. Finally, the proton affinity of poly(methylmetacrylate) oligomers is being discussed. Even though no absolute values for the proton affinity were found, the experimental and computational results help to understand the variation that accompanies the oligomers length. The second project presents the development an efficient and reproducible screening method for identifying low molecular weight compounds that bind to amyloid beta peptides (Abeta) peptides using electrospray ionization mass spectrometry (ESI-MS). Low molecular weight (LMW) compounds capable of interacting with soluble Abeta may be able to modulate/inhibit the Abeta aggregation process and serve as potential disease-modifying agents for Alzheimer’s disease. The present approach was used to rank the binding affinity of a library of compounds to Abeta1-40 peptide. The results obtained show that low molecular weight compounds bind similarly to Abeta1-42, Abeta1-40, as well as Abeta1-28 peptides and they underline the critical role of Abeta peptide charge motif in binding at physiological pH. Finally, some elements of structure-activity relationship (SAR) involved in the binding affinity of homotaurine to soluble Abeta peptides are discussed. As a third project, the gas phase binding of small molecules to the Abeta1-40 peptide generated by electrospray ionization has been explored with collision-induced dissociation mass spectrometry and kinetic rate theory. This project presents a simple procedure used to theoretically model the experimental breakdown diagrams for the Abeta1-40 peptide complexed with a series of aminosulfonate small molecules, namely homotaurine, 3-cyclohexylamino-2-hydroxy-1-propanesulfonic acid (CAPSO), 3-(1,3,4,9-tetrahydro-2H-beta-carbolin-2-yl) propane-1-sulfonic acid, 3-(1,3,4,9-tetrahydro-2H-beta-carbolin-2-yl)butane-1-sulfonic acid, and 3-(cyclohexylamino) propane-1-sulfonic acid. An alternative method employing an extrapolation procedure for the microcanonical rate constant, k(E), is also discussed.

As a first project, collision-induced dissociation experiments were carried out using electrospray ionisation mass spectrometry on gas phase complexes involving different poly(methylmetacrylate) oligomers with three amino acids: glycine, leucine, and phenylalanine. After acquiring breakdown diagrams, RRKM modeling was used to fit the experimental data in order to obtain the 0 K activation energy and the entropy of activation. These thermodynamic data were then used to understand the competing dissociation channels observed (except for gas phase complexes involving glycine that had only one dissociation channel). Molecular dynamics simulated annealing calculations were carried on the gas phase complexes to understand further the energetic and entropic effects involved as well as the 3D conformation of these complexes. Valuable insight information was found on the 3D conformations, on a qualitative level. Using rotational constants and vibrational harmonic frequencies, it was possible to evaluate the entropy variation between the experimentally observed competing channels. Reasonable agreement was found between the experimental and theoretical variations of entropies. Finally, the proton affinity of poly(methylmetacrylate) oligomers is being discussed. Even though no absolute values for the proton affinity were found, the experimental and computational results help to understand the variation that accompanies the oligomers length. The second project presents the development an efficient and reproducible screening method for identifying low molecular weight compounds that bind to amyloid beta peptides (Abeta) peptides using electrospray ionization mass spectrometry (ESI-MS). Low molecular weight (LMW) compounds capable of interacting with soluble Abeta may be able to modulate/inhibit the Abeta aggregation process and serve as potential disease-modifying agents for Alzheimer’s disease. The present approach was used to rank the binding affinity of a library of compounds to Abeta1-40 peptide. The results obtained show that low molecular weight compounds bind similarly to Abeta1-42, Abeta1-40, as well as Abeta1-28 peptides and they underline the critical role of Abeta peptide charge motif in binding at physiological pH. Finally, some elements of structure-activity relationship (SAR) involved in the binding affinity of homotaurine to soluble Abeta peptides are discussed. As a third project, the gas phase binding of small molecules to the Abeta1-40 peptide generated by electrospray ionization has been explored with collision-induced dissociation mass spectrometry and kinetic rate theory. This project presents a simple procedure used to theoretically model the experimental breakdown diagrams for the Abeta1-40 peptide complexed with a series of aminosulfonate small molecules, namely homotaurine, 3-cyclohexylamino-2-hydroxy-1-propanesulfonic acid (CAPSO), 3-(1,3,4,9-tetrahydro-2H-beta-carbolin-2-yl) propane-1-sulfonic acid, 3-(1,3,4,9-tetrahydro-2H-beta-carbolin-2-yl)butane-1-sulfonic acid, and 3-(cyclohexylamino) propane-1-sulfonic acid. An alternative method employing an extrapolation procedure for the microcanonical rate constant, k(E), is also discussed.

An Ab Initio/RRKM study of the reaction mechanism and product branching ratios of neutral-radical ethynyl (C2H) and cyano (CN) radical species with unsaturated hydrocarbons is performed. The reactions studied apply to cold conditions such as planetary atmospheres including Titan, the Interstellar Medium (ISM), icy bodies and molecular clouds. The reactions of C2H and CN additions to gaseous unsaturated hydrocarbons are an active area of study. NASA’s Cassini/Huygens mission found a high concentration of C2H and CN from photolysis of ethyne (C2H2) and hydrogen cyanide (HCN), respectively, in the organic haze layers of the atmosphere of Titan. The reactions involved in the atmospheric chemistry of Titan lead to a vast array of larger, more complex intermediates and products and may also serve as a chemical model of Earth’s primordial atmospheric conditions. The C2H and CN additions are rapid and exothermic, and often occur barrierlessly to various carbon sites of unsaturated hydrocarbons. The reaction mechanism is proposed on the basis of the resulting potential energy surface (PES) that includes all the possible intermediates and transition states that can occur, and all the products that lie on the surface. The B3LYP/6-311g(d,p) level of theory is employed to determine optimized electronic structures, moments of inertia, vibrational frequencies, and zero-point energy. They are followed by single point higher-level CCSD(T)/cc-vtz calculations, including extrapolations to complete basis sets (CBS) of the reactants and products. A microcanonical RRKM study predicts single-collision (zero-pressure limit) rate constants of all reaction paths on the potential energy surface, which is then used to compute the branching ratios of the products that result. These theoretical calculations are conducted either jointly or in parallel to experimental work to elucidate the chemical composition of Titan’s atmosphere, the ISM, and cold celestial bodies.

Cálculos usando métodos de estrutura eletrônica (Hartree-Fock, Mõller-Plesset de segunda ordem e DFT, B3LYP) foram efetuados pelo programa Gaussian98 em microcomputadores e estações de trabalho, com o objetivo de elucidar os canais de decomposição unimolecular de éter dietílico em fase gasosa e foram comparados com valores obtidos através de aproximações baseadas na mecânica estatística pela metodologia de Benson. O éter dietílico vem sendo usado como um aditivo para o óleo diesel e combustível alternativo e o conhecimento de suas vias vias de decomposição é fundamental nestas investigações. Dezesseis modos primários, do qual, quatro ocorrem através de cisão de ligações simples C-O, C-C,C(1)-H e C(2)-H e doze ocorrendo através de estados de transição cíclicos, os quais eliminam produtos como hidrogênio etano, acetaldeído, etano, álcool etílico, metil etil éter, metano, alguns carbenos e também di-radicais, foram considerados para a determinação das barreiras de ativação, entalpias de reação, entropias de reação e energia livre de Gibbs de reação. Vias primárias ocorrendo através de cisão de ligação, não reproduziram os valores experimentais para as barreiras de ativação, mas reproduziram de modo significante, valores da entalpia da reação. Eliminação de eteno e álcool etílico, ocorrendo através de um estado de transição de quatro centros, apresentou a barreira de ativação mais baixa. Acetaldeído e eliminação de etano, ocorrendo através de quatro centros, apresentou barreira significativamente alta, mas por outro lado, a menor entalpia, ligeiramente exotérmica por -0,8 kcal.mol-1. Eliminações 1,2 de metano e carbeno ocorrendo através de três centros, junto com a eliminação 1, 1 de hidrogênio e carbeno por três centros, eliminação 2,2 hidrogênio e carbeno por três centros e eliminação 1,4 de hidrogênio, acetaldeído e eteno por seis centros, apresentaram barreiras de ativação relativamente próximas mostrando que elas são competitivas entre si. Os valores computados dos canais que ocorrem por estados de transição cíclicos foram comparados com os resultados experimentais disponíveis e discutida a validade desta abordagem computacional para o estudo de reações unimoleculares de multi-canais. Determinação de parâmetros termoquímicos, como calor de formação para espécies radicalares dos canais de decomposição primário e alguns radicais alcóxidos, junto com a estimativa de afinidade eletrônica e protônica (com a abordagem ab initio Gaussian 2 a qual estima energias eletrônicas muito precisas) foram efetuados e seus resultados foram comparados com os valores experimentais disponíveis e valores obtidos através de energias de ligação e da regra de aditividade de Benson. / Calculations using methods of electronic structure(Hartree-Fock, second order Moller-Plesset and DFT: B3LYP) had been effected through the Gaussian98 program in microcomputers and workstations, with the objective to elucidate the unimolecular decomposition channels of diethyl ether in gaseous phase. These results also had been compared to those obtained by the methodology based in statistical mechanics through Benson\'s approach. Sixteen primary ways, which, four occur through the break of simple bonds C-O, C-C,C(1)-H, C(2)-H, and twelve occur through cyclical transistion states, which eliminate products as hydrogen, ethene, acetaldehyde, ethane, ethyl alcohol, methyl ethyl ether, methane, some carbenes and also diradicals. These products had been considered to the determination of the activation barriers, enthalpies of reaction, entropies of reaction and free energy of Gibbs of reaction. Primary ways occurring through the break of bonds, had not reproduced experimental values for the activation barriers, however they had reproduced in a significant way, values of the enthalpy of the reaction. Elimination of ethene and ethyl alcohol, occurring by a transition state of four centers, presented the lowest activation barrier. Acetaldehyde and ethane elimination occurring through four centers, presented a high significantly barrier, but on the other hand it presented the smallest enthalpy, lightly exothermic above -0,8 kcal.mol-1. Eliminations 1,2 of methane and carbene occurring through three centers, together with the elimination 1,1 of hydrogen and carbene through three centers, elimination 2,2 of hydrogen and carbene through three centers and elimination 1,4 of hydrogen, acetaldehyde and ethene through six centers, had presented relatively next activation barriers, showing that they are competitive among themselves. The computed values of the channels that occur through cyclical transition states had been compared with the available experimental results and the trustworthiness of this computational boarding for the study of unimolecular reactions in multi-channel had been dicussed. Determination of thermochemical parameters, as heat of formation for radicalar species of the primary channel of decomposition and some alcoxyde radicais, together with the estimative of the electronic and protonic affinities,( with the ab initio Gaussian 2 boarding which estimate very precise eletronic energies ) and their results had been compared with the available experimental values and with values gotten through energies of bond and Benson\'s additivity rule.

Cálculos usando métodos de estrutura eletrônica (Hartree-Fock, Mõller-Plesset de segunda ordem e DFT, B3LYP) foram efetuados pelo programa Gaussian98 em microcomputadores e estações de trabalho, com o objetivo de elucidar os canais de decomposição unimolecular de éter dietílico em fase gasosa e foram comparados com valores obtidos através de aproximações baseadas na mecânica estatística pela metodologia de Benson. O éter dietílico vem sendo usado como um aditivo para o óleo diesel e combustível alternativo e o conhecimento de suas vias vias de decomposição é fundamental nestas investigações. Dezesseis modos primários, do qual, quatro ocorrem através de cisão de ligações simples C-O, C-C,C(1)-H e C(2)-H e doze ocorrendo através de estados de transição cíclicos, os quais eliminam produtos como hidrogênio etano, acetaldeído, etano, álcool etílico, metil etil éter, metano, alguns carbenos e também di-radicais, foram considerados para a determinação das barreiras de ativação, entalpias de reação, entropias de reação e energia livre de Gibbs de reação. Vias primárias ocorrendo através de cisão de ligação, não reproduziram os valores experimentais para as barreiras de ativação, mas reproduziram de modo significante, valores da entalpia da reação. Eliminação de eteno e álcool etílico, ocorrendo através de um estado de transição de quatro centros, apresentou a barreira de ativação mais baixa. Acetaldeído e eliminação de etano, ocorrendo através de quatro centros, apresentou barreira significativamente alta, mas por outro lado, a menor entalpia, ligeiramente exotérmica por -0,8 kcal.mol-1. Eliminações 1,2 de metano e carbeno ocorrendo através de três centros, junto com a eliminação 1, 1 de hidrogênio e carbeno por três centros, eliminação 2,2 hidrogênio e carbeno por três centros e eliminação 1,4 de hidrogênio, acetaldeído e eteno por seis centros, apresentaram barreiras de ativação relativamente próximas mostrando que elas são competitivas entre si. Os valores computados dos canais que ocorrem por estados de transição cíclicos foram comparados com os resultados experimentais disponíveis e discutida a validade desta abordagem computacional para o estudo de reações unimoleculares de multi-canais. Determinação de parâmetros termoquímicos, como calor de formação para espécies radicalares dos canais de decomposição primário e alguns radicais alcóxidos, junto com a estimativa de afinidade eletrônica e protônica (com a abordagem ab initio Gaussian 2 a qual estima energias eletrônicas muito precisas) foram efetuados e seus resultados foram comparados com os valores experimentais disponíveis e valores obtidos através de energias de ligação e da regra de aditividade de Benson. / Calculations using methods of electronic structure(Hartree-Fock, second order Moller-Plesset and DFT: B3LYP) had been effected through the Gaussian98 program in microcomputers and workstations, with the objective to elucidate the unimolecular decomposition channels of diethyl ether in gaseous phase. These results also had been compared to those obtained by the methodology based in statistical mechanics through Benson\'s approach. Sixteen primary ways, which, four occur through the break of simple bonds C-O, C-C,C(1)-H, C(2)-H, and twelve occur through cyclical transistion states, which eliminate products as hydrogen, ethene, acetaldehyde, ethane, ethyl alcohol, methyl ethyl ether, methane, some carbenes and also diradicals. These products had been considered to the determination of the activation barriers, enthalpies of reaction, entropies of reaction and free energy of Gibbs of reaction. Primary ways occurring through the break of bonds, had not reproduced experimental values for the activation barriers, however they had reproduced in a significant way, values of the enthalpy of the reaction. Elimination of ethene and ethyl alcohol, occurring by a transition state of four centers, presented the lowest activation barrier. Acetaldehyde and ethane elimination occurring through four centers, presented a high significantly barrier, but on the other hand it presented the smallest enthalpy, lightly exothermic above -0,8 kcal.mol-1. Eliminations 1,2 of methane and carbene occurring through three centers, together with the elimination 1,1 of hydrogen and carbene through three centers, elimination 2,2 of hydrogen and carbene through three centers and elimination 1,4 of hydrogen, acetaldehyde and ethene through six centers, had presented relatively next activation barriers, showing that they are competitive among themselves. The computed values of the channels that occur through cyclical transition states had been compared with the available experimental results and the trustworthiness of this computational boarding for the study of unimolecular reactions in multi-channel had been dicussed. Determination of thermochemical parameters, as heat of formation for radicalar species of the primary channel of decomposition and some alcoxyde radicais, together with the estimative of the electronic and protonic affinities,( with the ab initio Gaussian 2 boarding which estimate very precise eletronic energies ) and their results had been compared with the available experimental values and with values gotten through energies of bond and Benson\'s additivity rule.

This research explores the links between the structure and dissociation energetics of ionized non-covalent complexes. In chapter 3, a large series of similar non-covalent complexes were probed using electrospray tandem mass spectrometry (ESI-MS/MS) and RRKM modelling in order to identify any trends in the dissociation energetics based on charge state, overall size of the complex, or size of the substrate. Ion mobility spectrometry (IMS) in conjunction with molecular mechanics/molecular dynamics (MM/MD) was used to study the conformations of these non-covalent complexes in order to determine if the same trends identified in the energetics could be corroborated independently based on structure. The system of study consisted of varying lengths of the synthetic polymer, polymethylmethacrylate (PMMA) complexed with singly or doubly protonated diaminoalkanes (DAA) of varying length. The critical energies of dissociation (E0) increased as the length of the polymer increased and was not significantly affected by the length of the singly protonated DAA substrates. The E0 of dissociation of doubly protonated complexes was strongly influenced by the length of the DAA; longer DAA substrates had greater separation of charge which decreased coulombic repulsion within the complex resulting in higher E0 values. MM/MD low energy structures of all complexes were validated with experimental IMS measurements and showed that the arrangement between the polymer and DAA were similar for different singly protonated DAAs. When doubly protonated, the length of DAA was the most important factor in determining the overall structure of the complex. In chapter 4, a direct link is shown between the observed E0 dissociation energies and the molecular conformations for eight different peptide–saccharide complexes containing either a tri-saccharide (d-(+)-raffinose and d-panose) or tetra-saccharide (stachyose and maltotetraose) with a small peptide (FLEEL and FLEEV). The E0 values were highly related to the overall conformation adopted by the non-covalent complex in the gas phase. Complexes containing peptide FLEE(L/V) with the tri-saccharide raffinose or panose had similar E0 of dissociation (∼0.64 eV) and similar conformations based on MM/MD simulations and IMS drift times. Conversely, for complexes containing a FLEE(L/V) peptide with one of the isomeric tetra-saccharides; stachyose had a E0 ∼0.08 eV greater than maltotetraose. This difference of intermolecular interaction was also reflected by the IMS drift times; maltotetraose in complex with FLEEV or FLEEL had a 5.9% and 2.3% faster IMS drift time than stachyose respectively. This indicated that the molecular arrangement between maltotetraose and the peptides was more compact than the stachyose-peptide complexes. In chapter 5, RRKM modelling of breakdown diagrams is not possible when the reactant ion signal is overlapped by other isobaric species. Trimeric, non-covalent complexes that contained two PMMA molecules and a doubly protonated DAA, [(PMMAa)(DAA+2H)(PMMAb)]+2, have m/z signals that contain multiple different complexes having the same total number of polymer repeat units but differ in the length of the each polymer. In this situation, the applicability of using the simple kinetic method to gain insight into relative binding energies was explored. The major factors which determined the suitability of the kinetic method for this system were identified as the structural arrangement of the reactant ion complex, possible reverse activation barriers, and the evaluations of Δ(ΔS‡). MM/MD simulations coupled with IMS suggests that within the reactant ion, the DAA is almost equally shared between two PMMA oligomers and that the two PMMA oligomers interact predominately with the DAA, and not with each other. MS/MS of the trimeric reactant complexes proceeds by neutral loss of one polymer and is suggested to proceed with little or no reverse activation barrier based on the low coulombic repulsion factors. The IMS drift times of [(PMMAa)(DAA+2H)]+2 complexes that were generated directly by ESI-MS or by dissociation of a trimeric, [(PMMAa)(DAA+2H)(PMMAb)]+2 complex were found to be identical. This provides some evidence that Δ(ΔS‡) ≈ Δ(ΔS) and using a statistical mechanics approach, Δ(ΔS) ≈ 0. The effective temperature (Teff) variable in the kinetic method expression was found to decrease as a function of the size of the trimeric complex, suggesting that the population distribution of the dissociating ensemble of complexes narrows as size increases. Overall, when RRKM fitting is not possible, the simple kinetic method could provide relative energetic ranking of competing dissociations reactions however the Teff term contributed to the greatest uncertainty in obtaining absolute quantities. Fitting MS/MS breakdown diagrams of non-covalent complexes with multiple dissociation channels is difficult due to the number of total fitting variables. Building from the simple kinetic method, chapter 6 shows that the relationship between the natural logarithm of competing fragment ions and reciprocal collision energy yields a branching relationship that allows for the sign of Δ(ΔS‡) and Δ(E0) between the channels to be obtained. Furthermore, the relationships between the fitting variables of RRKM modelling are empirically related to the theoretical branching relationship characteristics. This allowed for the fitting variables of all dissociation channels to be expressed as a function of a single channel so that the theoretical branching relationship matches the experimental branching relationship. Using this method, RRKM fitting of a MS/MS breakdown diagram for APCI ionized anthracene determined the E0 and ∆S‡ was 4.69 ± 0.29 eV and -3 ± 17 J K-1; 4.21 ±0.29 eV and -19 ±15 J K-1; and 4.81 ± 0.29 eV and 36 ±22 J K-1 for hydrogen loss, acetylene loss and diacetylene loss respectively. With one exception, these values are within experimental error of the iPEPICO derived energetic values. In chapter 7, MS/MS of ammoniated triacylglycerides at multiple collision energies and computational analysis are used to explain the cause of uneven dissociation rates of the FAs from different positions on the glycerol backbone. The loss of sn-1 and sn-3 FAs are found to have lower activation energies than the loss of the sn-2 position FA, however the loss of the FA from the sn-2 position is more entropically favourable. Theoretical MS/MS breakdown curves were fit to experimental values using RRKM theory to estimate the E0 of dissociation of FAs from the three glycerol positions. The E0 values for cleavage from the sn-1 and sn-3 positions were found to be approximately 1.52 eV, while that for the sn-2 position was highly dependent on the identity of the FA at that position. Computational structures and energy analysis suggest that an important step in the dissociation of [TAG+NH4]+ is the loss of ammonia. In a model system, glyceryl tributyrate, the loss of NH3 produced two distinct [TAG+H]+ product structures sitting 148 kJ and 160 kJ in energy above the ammoniated structure. The [TAG+H]+ structure that leads to the loss of the sn-1(3) is 12 kJ lower than the [TAG+H]+ structure that leads to the loss of the sn-2 FA. From this, the loss of a neutral FA that follows sits only an additional 35–48 kJ above the [TAG+H]+ structures. In Chapter 8, singly deprotonated β-cyclodextrin monomers, [(β-CD-H+]-1, and doubly deprotonated dimers, [(β-CD)2-2H+]-2, are both present following ESI-MS and have the same monoisotopic m/z. Similar to chapter 5, this makes it difficult to generate an MS/MS breakdown diagrams that can be modelled with RRKM theory. IMS was used to mobility separate [(β-CD-H+]-1 and [(β-CD)2-2H+]-2 and was followed by MS/MS of the [(β-CycD)2-2H+]-2 ion. A second problem when generating a MS/MS breakdown diagram of non-covalent complexes that contain identical components is that the fragment ions could have an identical monoisotopic m/z as the reactant ion. MS/MS of [(β-CycD)2-2H+]-2 results in two [(β-CD-H+]-1 fragments. To overcome this, breakdown diagrams were then generated by monitoring the changes in the isotopic profile. The RRKM derived E0 for dissociation of [(β-CycD)2-H+]-1 and [(β-CycD)2-2H+]-2 were 1.85 ± 0.11eV and 1.79 ± 0.09eV, respectively, corresponding to a slight decrease in complex stability due to increased charge-charge repulsion in the dianion.