D-Mannitol was simultaneously tritylated and acetylated to give 1,6-di-0-trityl-D-mannitol tetraacetate in 70% yield, using a procedure that gave a product completely free from triphenylcarbinol. Direct tritylation of D-mannitol yielded 1,6-di-O-trityl-D-mannitol which was crystallized, probably as a solvated molecule, from alcohol and water. Acetylation of this compound gave 1,6-di-0-trityl-D-mannitol tetraacetate in good yield. Detritylation of 1,6-di-O-trityl-D-mannitol tetraacetate to D-mannitol-2,3,4,5-tetraacetate was effected by hydrogen bromide in glacial acetic acid, catalytic hydrogenolysis over platinum oxide, and refluxing with dilute acetic acid, and the latter convenient method was found to give the highest yield of pure product. A sirupy di-O-methyl-D-mannitol was prepared from D-raannitol-2,3,4,5-tetraacetate by methylation followed by deacetylation on an ion exchange column. The structure of is this compound has not been established. A nearly colorless, amorphous solid was isolated in 2.7% yield by 50% aqueous ethanol extraction of the finely-divided wood of a Western Hemlock tree (Tsuga heterophylla) which had previously been exhaustively extracted with benzene. Treatment of this material with concentrated hydrochloric acid in methanol yielded a purple solid whose red-violet methanol solution was stable to ordinary light. The paper chromatographic behaviour and color reactions compared with a red rose petal extract and the methoxyl content indicated that the purple solid was probably an anthocyanidin containing methoxyl groups. Acetylation of the original solid extract with pyridine and acetic anhydride gave a dextrorotatory yellow sirup soluble in chloroform. / Science, Faculty of / Chemistry, Department of / Graduate

L’ANS recombinante de Vitis vinifera (VvANS) a été exprimée chez E. coli, et purifiée par chromatographie d’affinité sur colonne Nickel. La production et purification de l'holoenzyme chargée en fer a été mise en point, afin d’éviter les réactions d'oxydation non-enzymatique incontrôlée en présence de sel de Fe(II). Un complexe VvANS-Fe(II) stable est formé en présence d'α- cétoglutarate et d'ascorbate, complexe qui est catalytiquement actif à PO2 ambiante en l'absence de sel de Fe(II). La transformation de la (+)-catéchine par VvANS a été étudiée avec ou sans ascorbate, en utilisant le complexe VvANSFe( II) ou l’holoenzyme co-incubée en présence de sulfate ferreux, afin d’étudier le rôle de l’ascorbate. Aucune activité enzymatique n’a été observée en l'absence d’ascorbate, ce qui indique qu'il s'agit d'un cofacteur indispensable de VvANS. Un adduit covalent ascorbate-cyanidine est produit in vitro, mais seulement en l'absence d'un autre réducteur nucléophile majeur, le glutathion GSH. Les deux stéréoisomères de la leucocyanidine (flavan-diols 3,4-cis et 3,4-trans), substrats potentiels de VvANS, mais non commerciaux, ont été synthétisés par réduction de la dihydroquercétine par NaBH4, puis identifiés par RMN du proton. L'analyse des deux stéréoisomères par spectrométrie de masse en tandem (MS/MS) montre que leurs voies de fragmentation MS/MS sont distinctes et peuvent être utilisées pour les distinguer lors de leur production. Les deux stéréoisomères sont stables en milieu aqueux congelé à -20°C. Douze flavonoïdes de quatre familles distinctes (flavanones, dihydroflavonols, flavan-3-ols et flavan-3,4-diols) ont été testés comme substrats potentiels. Tous les produits enzymatiques ont été purifiés par HPLC en phase inverse, puis identifiés par MS/MS, avec les résultats suivants: 1) Seuls les dihydroflavonols de configuration (2R,3R) sont acceptés comme substrats par VvANS dont l'activité diminue avec le nombre de groupements hydroxyles du cycle B. 2) Seuls les flavan-3-ols ou flavan-3,4-diols de configuration (2R,3S) ayant un catéchol ou trois OH phénoliques vicinaux sur le cycle B sont acceptés comme substrats. 3) La naringénine n'est pas substrat de VvANS, sans doute en raison de l'absence de groupement hydroxyle en C3. […] Le glutathion GSH est un puissant nucléophile, réducteur et piégeur de radicaux libres, qui est abondant dans la baie de raisin. Nous avons donc étudié son effet sur l'activité de VvANS avec tous les substrats identifiés. GSH n’a pas d'effet sur la transformation des dihydroflavonols et des flavan-3,4-diols, mais il modifie considérablement le mode de transformation de la (+)-catéchine et de la (+)-gallocatéchine. En présence de (+)-catéchine et de GSH, on observe deux produits majeurs, la cyanidine et un adduit thioéther cyanidine-glutathion, et le rendement de production est beaucoup plus élevé qu'en l'absence de GSH. De plus, l’adduit covalent ascorbate-cyanidine et le dimère issu de la (+)-catéchine obtenus lors de la réaction réalisée en l'absence de GSH ont disparu. Nos données suggèrent que l'adduit covalent cyanidine-glutathion est un thioéther en C4 qui fait l'objet d'un équilibre de tautomérisation céto-énolique en C3, et se décompose en cyanidine et GSH. En présence de (+)-gallocatéchine, un adduit thioéther similaire delphinidine-glutathion est aussi observé. Pour tester l'éventuelle spécificité de GSH, trois autres mercaptans (thiomalate, cystéine et cystéamine) ont été testés et aucun adduit similaire n’a été observé, ce qui suggère que GSH est un ligand spécifique, et pourrait être un coenzyme de VvANS. Nos résultats suggèrent que les anthocyanidines pourraient être produites in vivo à partir d'un substrat flavan-3-ol (catéchine ou gallocatéchine) via un intermédiaire thioéther de glutathion, alors que le stéréoisomère naturel (3,4-cis) de la leucocyanidine n'est pas transformé en cyanidine. / Recombinant anthocyanidin synthase from Vitis vinifera (VvANS) has been expressed in E. coli, and purified by nickel affinity chromatography. The production and purification of the iron-loaded enzyme has been developed in order to avoid uncontrolled nonenzymatic oxidation reactions in the presence of Fe(II) salt. A stable VvANS-Fe(II) complex is formed in the presence of 2-oxoglutarate and ascorbate, and this complex is catalytically active at ambient PO2 in the absence of Fe(II) salt. The transformation of (+)-catechin by VvANS has been studied with and without ascorbate, by using either the VvANSFe( II) complex or the holoenzyme co-incubated with ferrous sulfate, to investigate the role of ascorbate. No enzyme activity has been observed in the absence of ascorbate, which means that it is an essential enzyme cofactor. A covalent adduct ascorbate-cyanidin is produced in vitro, but only in the absence of glutathione (GSH), another major nucleophilic and reducing agent. The two stereoisomers of leucocyanidin (3,4-cis et 3,4-trans flavan-diols) which were expected to behave as substrates of VvANS, are not commercial and were synthesized by reduction of dihydroquercetin by NaBH4, and characterized by proton NMR. The analysis of the two stereoisomers by means of tandem mass spectrometry (MS/MS) shows that their fragmentation pathways are distinct and may be used to distinguish them during their production. The two stereoisomers are stable in frozen aqueous medium at -20°C. Twelve flavonoids of four distinct families (flavanones, dihydroflavonols, flavan-3-ols et flavan-3,4-diols) were tested as potential substrates of VvANS. All enzymatic products were purified by means of reverse-phase HPLC and characterized by MS/MS, with the following results: 1) Only dihydroflavonols of (2R,3R) configuration are accepted as substrates by VvANS, the activity decreasing with the number of hydroxyl groups of ring B. 2) Only flavan-3-ols or flavan-3,4-diols of (2R,3S) configuration having either a catechol or three vicinal phenolic OH on ring B are accepted as substrates. 3) Naringenin is not substrate of VvANS, most likely because a C3 hydroxyl group is missing. […] Glutathione GSH is a powerful nucleophilic and reducing agent as well as a free radical scavenger, which is abundant in grape berries. We therefore studied its effect on VvANS activity with all identified substrates. GSH has no effect on the transformation of dihydroflavonols and flavan-3,4-diols, but it considerably modifies the transformation pattern of (+)- catechin and (+)-gallocatechin. In the presence of (+)-catechin and GSH, we observe two major products, cyanidin and a cyanidin-glutathione thioether, with production yields which are much higher than in the absence of GSH. Moreover, the ascorbate-cyanidin covalent adduct and the (+)-catechin dimer that had been obtained in the absence of GSH have disappeared. Our data suggest that the cyanidin-glutathione adduct is a C4-thioether which is in equilibrium between the two keto-enolic tautomeric forms at C3, and decomposes into cyanidin and GSH. In the presence of (+)-gallocatechin, a similar delphinidin-glutathione thioether adduct is also observed. In order to test the possible specificity of GSH as a cofactor, three other mercaptans (thiomalate, cysteine and cysteamine) were tested, and no similar product was observed, which suggests that GSH is a specific ligand, and might be a coenzyme of VvANS. Our results suggest that anthocyanidins could be produced in vivo from a flavan-3-ol substrate (catechin or gallocatechin) via a glutathione thioether intermediate, whereas the natural 3,4-cis stereoisomer of leucocyanidin is not transformed into cyanidin by VvANS.

Grazing cattle on alfalfa (Medicago sativa L.) would be economically beneficial, but its rapid initial rate of protein degradation results in pasture bloat, low efficiency of protein utilization and excessive N pollution into the environment. Introducing a gene that stimulates the accumulation of mono/polymeric anthocyanidins might reduce the ruminal protein degradation rate and reduce bloat related foam stability. The overall objective of this thesis was to evaluate newly developed anthocyanidin-accumulating Lc-alfalfa progeny for nutritional properties (composition, site of degradation and molecular structure), environmental emissions and bloat characteristics. The objective of the first study was to determine survival and phytochemical and chemical profiles of Lc-alfalfa progeny (BeavLc1, RambLc3 and RangLc4) and their non-transgenic (NT) parental cultivars (Beaver, Rambler and Rangelander). Lc-alfalfa forage accumulated enhanced amounts of anthocyanidin, with an average concentration of 197.4 µg/g DM, while proanthocyanidin (i.e. condensed tannins) were not detected. Both of these metabolites were absent in the NT-parental varieties. Lc-alfalfa progeny had ~3 % less crude protein (CP) and ~3 % more carbohydrates (CHO), which resulted in their 11 g/kg lower N:CHO ratio compared with NT-alfalfa. Total rumen-degradable N:CHO ratio based on chemical analysis was 12.9 g/kg lower in Lc-alfalfa compared with NT-alfalfa. The objective of the second study was to evaluate in vitro degradation, fermentation and microbial-N partitioning of three forage color phenotypes [green, light purple-green (LPG) and purple-green (PG)] within Lc-progeny and their parental green NT-alfalfa varieties. Purple-green-Lc alfalfa accumulated more anthocyanidin than Green-Lc with LPG-Lc intermediate. Gas, methane and ammonia accumulation rates were slower for the two purple-Lc phenotypes compared with NT-alfalfa with Green-Lc intermediate. Effective degradable DM and N were lower in the three Lc-phenotypes compared with NT-alfalfa. Anthocyanidin concentration correlated negatively with gas and methane production rates and effective degradability of DM and N. The objectives of the third study were to evaluate in situ ruminal degradation characteristics and synchronization ratios, and to model protein availability to dairy cattle and net energy for lactation of three Lc-alfalfa progenies, BeavLc1, RambLc3 and RangLc4 and the cultivar AC Grazeland (selected for a low initial rate of ruminal degradation). Anthocyanidin accumulation was on average 163.4 ìg/g DM in the three Lc-progeny while AC Grazeland did not accumulate anthocyanidin. The basic chemical composition of the original samples, soluble and potentially degradable fractions and degradation characteristics of crude protein and carbohydrates were similar in Lc-alfalfa and AC Grazeland. The undegradable in situ crude protein and neutral detergent fiber fraction were, respectively, 1.3 %CP and 4.8 %CHO lower in the three Lc-progeny compared with AC Grazeland. Lc-alfalfa had a 0.34 MJ/kg DM higher net energy for lactation and tended to have a 11.9, 6.9 and 8.4 g/kg DM higher rumen degradable protein, rumen degraded protein balance and intestinal available protein, respectively, compared with AC Grazeland,. The hourly rumen degraded protein balance included an initial and substantial peak (over-supply) of protein relative to energy which was highest in RangLc4 and lowest in RambLc3. The hourly rumen degraded protein balance between 4 and 24 h was similar and more balanced for all four alfalfa populations. The objective of the fourth study was to determine foam formation and stability in vitro from aqueous leaf extracts of three Lc-alfalfa progeny (BeavLc1, RambLc3, RangLc4), parental NT-alfalfa and AC Grazeland (bloat reduced cultivar) harvested in the field at 07:00 or 18:00 h. Anthocyanidin accumulation averaged 247.5 ìg/g DM in the leaves of the three Lc-progeny. There was an interaction between population and harvest time for the foam parameters. Initial foam volume (0 min) and final foam volume (150 min) at 07:00 h were lower for AC Grazeland compared with all other treatments and lower for RangLc4 compared with the other two Lc-progeny at 0 min and NT-alfalfa at 150 min; while from the 18:00 h harvest, initial foam volume was larger for NT-alfalfa and final foam volume was larger for RambLc3 compared with AC Grazeland, BeavLc1 and RangLc4. Foam formation correlated positively (R = 0.30 to 0.44) with leaf DM content, leaf extract protein and ethanol-film content, spectroscopic vibration intensity due to all carbohydrates (CHOVI) and amide I:amide II ratio and negatively (R = -0.33 and -0.34; P<0.05) with á-helix:â-sheet ratio and amide I:CHOVI. Final foam volume correlated negatively (R = -0.53 to -0.25; P<0.05) with leaf extract pH, spectroscopic vibration intensity due to all protein structures, structural carbohydrates (SCVI) and lipids (CH2 and CH3 asymmetric stretching) and amide I:CHOVI ratio and corelated positively (R = 0.39 to 0.44; P<0.05) with CHOVI, amideI:SCVI ratio and CHOVI:SCVI ratio. In conclusion, all Lc-alfalfa progeny and phenotypes accumulated anthocyanidin in their forage. Lc-alfalfa progeny had lower protein and higher carbohydrate content which improved the nitrogen to carbohydrate balance compared to their parental NT-alfalfa cultivars. Rate of fermentation and effective degradability in vitro reduced for both purple anthocyanidin-accumulating Lc-alfalfa phenotypes compared with NT-alfalfa. Intestinal protein availability tended to be higher and net energy for lactation was higher from Lc-alfalfa progeny for dairy cattle compared with AC Grazeland. Foaming properties were reduced in Lc-alfalfa progeny compared with parental non-transgenic alfalfa but not compared with AC Grazeland. However, differences between the Lc-alfalfa progeny and other cultivars were small. Therefore, further increases in mono/polymeric anthocyanidin accumulation in alfalfa are required in order to develop an alfalfa cultivar with superior nutritional and bloat preventing characteristics compared to currently available alfalfa cultivars.

Grazing cattle on alfalfa (Medicago sativa L.) would be economically beneficial, but its rapid initial rate of protein degradation results in pasture bloat, low efficiency of protein utilization and excessive N pollution into the environment. Introducing a gene that stimulates the accumulation of mono/polymeric anthocyanidins might reduce the ruminal protein degradation rate and reduce bloat related foam stability. The overall objective of this thesis was to evaluate newly developed anthocyanidin-accumulating Lc-alfalfa progeny for nutritional properties (composition, site of degradation and molecular structure), environmental emissions and bloat characteristics. The objective of the first study was to determine survival and phytochemical and chemical profiles of Lc-alfalfa progeny (BeavLc1, RambLc3 and RangLc4) and their non-transgenic (NT) parental cultivars (Beaver, Rambler and Rangelander). Lc-alfalfa forage accumulated enhanced amounts of anthocyanidin, with an average concentration of 197.4 µg/g DM, while proanthocyanidin (i.e. condensed tannins) were not detected. Both of these metabolites were absent in the NT-parental varieties. Lc-alfalfa progeny had ~3 % less crude protein (CP) and ~3 % more carbohydrates (CHO), which resulted in their 11 g/kg lower N:CHO ratio compared with NT-alfalfa. Total rumen-degradable N:CHO ratio based on chemical analysis was 12.9 g/kg lower in Lc-alfalfa compared with NT-alfalfa. The objective of the second study was to evaluate in vitro degradation, fermentation and microbial-N partitioning of three forage color phenotypes [green, light purple-green (LPG) and purple-green (PG)] within Lc-progeny and their parental green NT-alfalfa varieties. Purple-green-Lc alfalfa accumulated more anthocyanidin than Green-Lc with LPG-Lc intermediate. Gas, methane and ammonia accumulation rates were slower for the two purple-Lc phenotypes compared with NT-alfalfa with Green-Lc intermediate. Effective degradable DM and N were lower in the three Lc-phenotypes compared with NT-alfalfa. Anthocyanidin concentration correlated negatively with gas and methane production rates and effective degradability of DM and N. The objectives of the third study were to evaluate in situ ruminal degradation characteristics and synchronization ratios, and to model protein availability to dairy cattle and net energy for lactation of three Lc-alfalfa progenies, BeavLc1, RambLc3 and RangLc4 and the cultivar AC Grazeland (selected for a low initial rate of ruminal degradation). Anthocyanidin accumulation was on average 163.4 ìg/g DM in the three Lc-progeny while AC Grazeland did not accumulate anthocyanidin. The basic chemical composition of the original samples, soluble and potentially degradable fractions and degradation characteristics of crude protein and carbohydrates were similar in Lc-alfalfa and AC Grazeland. The undegradable in situ crude protein and neutral detergent fiber fraction were, respectively, 1.3 %CP and 4.8 %CHO lower in the three Lc-progeny compared with AC Grazeland. Lc-alfalfa had a 0.34 MJ/kg DM higher net energy for lactation and tended to have a 11.9, 6.9 and 8.4 g/kg DM higher rumen degradable protein, rumen degraded protein balance and intestinal available protein, respectively, compared with AC Grazeland,. The hourly rumen degraded protein balance included an initial and substantial peak (over-supply) of protein relative to energy which was highest in RangLc4 and lowest in RambLc3. The hourly rumen degraded protein balance between 4 and 24 h was similar and more balanced for all four alfalfa populations. The objective of the fourth study was to determine foam formation and stability in vitro from aqueous leaf extracts of three Lc-alfalfa progeny (BeavLc1, RambLc3, RangLc4), parental NT-alfalfa and AC Grazeland (bloat reduced cultivar) harvested in the field at 07:00 or 18:00 h. Anthocyanidin accumulation averaged 247.5 ìg/g DM in the leaves of the three Lc-progeny. There was an interaction between population and harvest time for the foam parameters. Initial foam volume (0 min) and final foam volume (150 min) at 07:00 h were lower for AC Grazeland compared with all other treatments and lower for RangLc4 compared with the other two Lc-progeny at 0 min and NT-alfalfa at 150 min; while from the 18:00 h harvest, initial foam volume was larger for NT-alfalfa and final foam volume was larger for RambLc3 compared with AC Grazeland, BeavLc1 and RangLc4. Foam formation correlated positively (R = 0.30 to 0.44) with leaf DM content, leaf extract protein and ethanol-film content, spectroscopic vibration intensity due to all carbohydrates (CHOVI) and amide I:amide II ratio and negatively (R = -0.33 and -0.34; P<0.05) with á-helix:â-sheet ratio and amide I:CHOVI. Final foam volume correlated negatively (R = -0.53 to -0.25; P<0.05) with leaf extract pH, spectroscopic vibration intensity due to all protein structures, structural carbohydrates (SCVI) and lipids (CH2 and CH3 asymmetric stretching) and amide I:CHOVI ratio and corelated positively (R = 0.39 to 0.44; P<0.05) with CHOVI, amideI:SCVI ratio and CHOVI:SCVI ratio. In conclusion, all Lc-alfalfa progeny and phenotypes accumulated anthocyanidin in their forage. Lc-alfalfa progeny had lower protein and higher carbohydrate content which improved the nitrogen to carbohydrate balance compared to their parental NT-alfalfa cultivars. Rate of fermentation and effective degradability in vitro reduced for both purple anthocyanidin-accumulating Lc-alfalfa phenotypes compared with NT-alfalfa. Intestinal protein availability tended to be higher and net energy for lactation was higher from Lc-alfalfa progeny for dairy cattle compared with AC Grazeland. Foaming properties were reduced in Lc-alfalfa progeny compared with parental non-transgenic alfalfa but not compared with AC Grazeland. However, differences between the Lc-alfalfa progeny and other cultivars were small. Therefore, further increases in mono/polymeric anthocyanidin accumulation in alfalfa are required in order to develop an alfalfa cultivar with superior nutritional and bloat preventing characteristics compared to currently available alfalfa cultivars.

Philosophiae Doctor - PhD (Biotechnology) / The 'Bon Rouge' pear cultivar was developed from a bud mutation on a 'Bon Chretien' pear tree. The latter is characterised by green fruit skin and leaves, while 'Bon Rouge' is characterised by red leaves and red fruit skin as a result of the production of anthocyanin and other pigments. Branch forming buds on 'Bon Rouge' trees often revert to the parent phenotype producing green leaves and fruit skin. The occurrence of both phenotypes on the same tree presents a unique model to study gene expression associated with anthocyanin production in a similar genetic background under the same set of environmental condition.

A natureza das ligações de hidrogênio e a influência destas interações na estrutura eletrônica de complexos neutros, catiônicos, aniônicos e radicalares foi estudada utilizando-se análises geométricas, energéticas, eletrônicas e topológicas. Inicialmente, verificaram-se alterações na aromaticidade do cátion pirílio após a complexação com uma a três moléculas de água. Tais complexos foram ainda estudados em meio reacional com constante dielétrica igual a da água com o emprego do modelo PCM (Polarizable Continuum Model). Adicionalmente, os efeitos da hidroxilação na estrutura eletrônica dos cátions benzopirílio e flavílio foram considerados. Posteriormente, analisaram-se os efeitos das fortes ligações de hidrogênio na ressonância do grupo carboxila em complexos formados entre o radical hidroperoxil e os ácidos fórmico, acético e trifluoroacético. Como extensão desse trabalho, estudos envolvendo complexos obtidos com e sem restrições na otimização de geometria possibilitaram obter informações a respeito da ressonância dos grupos carboxila e carboxilato quando o fluoreto de hidrogênio interage linear ou perpendicularmente com todos os átomos do ácido fórmico e do ânion formiato. O desenvolvimento das atividades supracitadas compreendeu a análise da função de onda pelos métodos NBO (Natural Bond Orbital), NSA (Natural Steric Analysis), NRT (Natural Resonance Theory) e AIM (Atoms in Molecules). As alterações em parâmetros geométricos e nas cargas atômicas foram consideradas. Uma análise energética foi realizada com o emprego do método de decomposição de energia proposto por Xantheas. As freqüências vibracionais e a intensidade das bandas do estiramento do grupo X-H, doador da ligação de hidrogênio, foram analisadas. As densidades de spin para os complexos radicalares também foram obtidas. A influência das ligações de hidrogênio e o efeito do substituinte na aromaticidade dos cátions foram verificados com o emprego dos métodos e índices NICS (Nucleus Independent Chemical Shifts), HOMA (Harmonic Oscillator Model of Aromaticity), HOSE (Harmonic Oscillator Stabilization Energy) e PDI (para-Delocalization Index). Os cálculos foram efetuados com os modelos B3LYP/6-31+G(d,p), B3LYP/6-311++G(3df,3pd) e UB3LYP/6-311++G(3df,3pd). Ocasionalmente, outras funções de base (EPR-III e cc-pVDZ), assim como o método MP2, foram utilizados para testar a confiabilidade dos resultados obtidos. As interações intermoleculares pouco alteraram a estrutura eletrônica e a aromaticidade do cátion pirílio. Analogamente, a substituição de um átomo de hidrogênio por um grupo hidroxila em diversas posições dos cátions benzopirílio e flavílio também não provocou modificações muito significativas na estrutura eletrônica desses cátions, embora tenha se verificado uma dependência da aromaticidade com a posição da hidroxila. Por outro lado, a distorção geométrica associada às ligações de hidrogênio foram responsáveis pelo incremento ou diminuição da ressonância do grupo carboxila nos ácidos fórmico, acético, trifluoroacético e do grupo carboxilato no ânion formiato. Os efeitos dos grupos doador e sacador de elétrons na estabilização dos complexos radicalares foram evidenciados. Adicionalmente, pode-se atribuir um caráter covalente parcial em algumas ligações de hidrogênio. / The nature of hydrogen bonds and their influence on electronic structure of neutral, cationic, anionic, and radical complexes was studied by using geometric, energetic, electronic, and topological analysis. The changes in aromaticity of the pyrylium cation upon complexation with one up to three water molecules were investigated. The PCM (Polarizable Continuum Model) model was employed to study the pyrylium-water complexes in a water reaction medium. In addition, the effects of hydroxylation on electronic structure of the benzopyrylium and flavilium cations were also considered. In addition, the effects of strong hydrogen bonds on carboxyl group resonance in the complexes formed between the hydroperoxyl radical and formic, acetic, and trifluoroacetic acids were analyzed. In extension of this work, studies including complexes, obtained with and without geometric restrictions, provided information about the resonance of the carboxyl and carboxylate groups when the hydrogen fluoride interacts, linear or perpendicularly, with all atoms of formic acid and formate anion. The analysis of the wavefunction by using NBO (Natural Bond Orbital), NSA (Natural Steric Analysis), NRT (Natural Resonance Theory), and AIM (Atoms in Molecules) methods was necessary to the development of the above mentioned activities. The changes in geometric parameters and atomic charges were also considered. An energetic analysis of complexes was done with the energy decomposition method proposed by Xantheas. The vibrational frequencies and the intensity of the X-H (hydrogen bond donor group) stretching bands were studied. The spin densities for the radical complexes were also obtained. The Nucleus Independent Chemical Shifts (NICS), Harmonic Oscillator Model of Aromaticity (HOMA), HOSE (Harmonic Oscillator Stabilization Energy), and PDI (para-Delocalization Index) aromaticity criteria were employed to verify the hydrogen bond influence and the effect of hydroxylation in the aromaticity of the cations. The calculations were carried out by using B3LYP/6-31+G(d,p), B3LYP/6-311++G(3df,3pd), and UB3LYP/6-311++G(3df,3pd) models. Occasionally, other basis set (EPR-III and cc-pVDZ), as well as the MP2 method, were applied to test the accuracy of the results. The intermolecular interactions lead to small alterations in the electronic structure and aromaticity of pyrylium cation. Similarly, the substitution at different positions of the benzopyrylium and flavilium cations by a hydroxyl group does not cause significant changes in the electronic structure of these cations. However, a dependence of the hydroxyl group position on aromaticity was observed. On the other hand, for formic, acetic, trifluoroacetic acids, as well as for the formate anion, the resonance of the carboxyl and carboxylate groups is affected not only by the geometric distortions but also by the hydrogen bonds. The effects of the electron-donating and electron-withdrawing groups in the stabilization of radical complexes were characterized. Furthermore, a partial covalent character can be attributed to some hydrogen bonds.

A natureza das ligações de hidrogênio e a influência destas interações na estrutura eletrônica de complexos neutros, catiônicos, aniônicos e radicalares foi estudada utilizando-se análises geométricas, energéticas, eletrônicas e topológicas. Inicialmente, verificaram-se alterações na aromaticidade do cátion pirílio após a complexação com uma a três moléculas de água. Tais complexos foram ainda estudados em meio reacional com constante dielétrica igual a da água com o emprego do modelo PCM (Polarizable Continuum Model). Adicionalmente, os efeitos da hidroxilação na estrutura eletrônica dos cátions benzopirílio e flavílio foram considerados. Posteriormente, analisaram-se os efeitos das fortes ligações de hidrogênio na ressonância do grupo carboxila em complexos formados entre o radical hidroperoxil e os ácidos fórmico, acético e trifluoroacético. Como extensão desse trabalho, estudos envolvendo complexos obtidos com e sem restrições na otimização de geometria possibilitaram obter informações a respeito da ressonância dos grupos carboxila e carboxilato quando o fluoreto de hidrogênio interage linear ou perpendicularmente com todos os átomos do ácido fórmico e do ânion formiato. O desenvolvimento das atividades supracitadas compreendeu a análise da função de onda pelos métodos NBO (Natural Bond Orbital), NSA (Natural Steric Analysis), NRT (Natural Resonance Theory) e AIM (Atoms in Molecules). As alterações em parâmetros geométricos e nas cargas atômicas foram consideradas. Uma análise energética foi realizada com o emprego do método de decomposição de energia proposto por Xantheas. As freqüências vibracionais e a intensidade das bandas do estiramento do grupo X-H, doador da ligação de hidrogênio, foram analisadas. As densidades de spin para os complexos radicalares também foram obtidas. A influência das ligações de hidrogênio e o efeito do substituinte na aromaticidade dos cátions foram verificados com o emprego dos métodos e índices NICS (Nucleus Independent Chemical Shifts), HOMA (Harmonic Oscillator Model of Aromaticity), HOSE (Harmonic Oscillator Stabilization Energy) e PDI (para-Delocalization Index). Os cálculos foram efetuados com os modelos B3LYP/6-31+G(d,p), B3LYP/6-311++G(3df,3pd) e UB3LYP/6-311++G(3df,3pd). Ocasionalmente, outras funções de base (EPR-III e cc-pVDZ), assim como o método MP2, foram utilizados para testar a confiabilidade dos resultados obtidos. As interações intermoleculares pouco alteraram a estrutura eletrônica e a aromaticidade do cátion pirílio. Analogamente, a substituição de um átomo de hidrogênio por um grupo hidroxila em diversas posições dos cátions benzopirílio e flavílio também não provocou modificações muito significativas na estrutura eletrônica desses cátions, embora tenha se verificado uma dependência da aromaticidade com a posição da hidroxila. Por outro lado, a distorção geométrica associada às ligações de hidrogênio foram responsáveis pelo incremento ou diminuição da ressonância do grupo carboxila nos ácidos fórmico, acético, trifluoroacético e do grupo carboxilato no ânion formiato. Os efeitos dos grupos doador e sacador de elétrons na estabilização dos complexos radicalares foram evidenciados. Adicionalmente, pode-se atribuir um caráter covalente parcial em algumas ligações de hidrogênio. / The nature of hydrogen bonds and their influence on electronic structure of neutral, cationic, anionic, and radical complexes was studied by using geometric, energetic, electronic, and topological analysis. The changes in aromaticity of the pyrylium cation upon complexation with one up to three water molecules were investigated. The PCM (Polarizable Continuum Model) model was employed to study the pyrylium-water complexes in a water reaction medium. In addition, the effects of hydroxylation on electronic structure of the benzopyrylium and flavilium cations were also considered. In addition, the effects of strong hydrogen bonds on carboxyl group resonance in the complexes formed between the hydroperoxyl radical and formic, acetic, and trifluoroacetic acids were analyzed. In extension of this work, studies including complexes, obtained with and without geometric restrictions, provided information about the resonance of the carboxyl and carboxylate groups when the hydrogen fluoride interacts, linear or perpendicularly, with all atoms of formic acid and formate anion. The analysis of the wavefunction by using NBO (Natural Bond Orbital), NSA (Natural Steric Analysis), NRT (Natural Resonance Theory), and AIM (Atoms in Molecules) methods was necessary to the development of the above mentioned activities. The changes in geometric parameters and atomic charges were also considered. An energetic analysis of complexes was done with the energy decomposition method proposed by Xantheas. The vibrational frequencies and the intensity of the X-H (hydrogen bond donor group) stretching bands were studied. The spin densities for the radical complexes were also obtained. The Nucleus Independent Chemical Shifts (NICS), Harmonic Oscillator Model of Aromaticity (HOMA), HOSE (Harmonic Oscillator Stabilization Energy), and PDI (para-Delocalization Index) aromaticity criteria were employed to verify the hydrogen bond influence and the effect of hydroxylation in the aromaticity of the cations. The calculations were carried out by using B3LYP/6-31+G(d,p), B3LYP/6-311++G(3df,3pd), and UB3LYP/6-311++G(3df,3pd) models. Occasionally, other basis set (EPR-III and cc-pVDZ), as well as the MP2 method, were applied to test the accuracy of the results. The intermolecular interactions lead to small alterations in the electronic structure and aromaticity of pyrylium cation. Similarly, the substitution at different positions of the benzopyrylium and flavilium cations by a hydroxyl group does not cause significant changes in the electronic structure of these cations. However, a dependence of the hydroxyl group position on aromaticity was observed. On the other hand, for formic, acetic, trifluoroacetic acids, as well as for the formate anion, the resonance of the carboxyl and carboxylate groups is affected not only by the geometric distortions but also by the hydrogen bonds. The effects of the electron-donating and electron-withdrawing groups in the stabilization of radical complexes were characterized. Furthermore, a partial covalent character can be attributed to some hydrogen bonds.