INFORME FINAL"Alguns importants descobriments de la ciència són accidentals. Això fou, certament, el cas del ful·lerè C60. Quan ara fa quatre anys, l'any 2000, vaig iniciar el treball d'investigació sobre els metal·loful·lerens m'havien arribat veus de la importància mediàtica d'aquestes noves estructures però no de la intensa història del seu descobriment. Aquesta m'ha fascinat. Sobretot quan l'any 2003 vaig tenir la possibilitat de fer una estada al grup del professor Kroto de la universitat de Sussex (Gran Bretanya). Llavors els textos, sorprenents però inerts, que havia llegit sobre la història del descobriment dels ful·lerens prenien forma en espais coneguts i personatges propers". Aquest és el prefaci de la meva tesi. Segurament aquesta estada a Sussex al grup d'un dels descobridors dels ful·lerens ha estat el moment més entranyable en aquesta carrera científica per l'estudi dels metal·loful·lerens.A partir de la síntesi de la molècula de C60 en quantitats apreciables la química dels ful·lerens en general, i en particular amb metalls de transició, ha experimentat un increment espectacular. L'objectiu d'aquesta tesi és l'estudi de l'estructura i la reactivitat de diferents derivats organometàl·lics de ful·lerens o metal·loful·lerens. El treball d'investigació usa les eines de la química computacional (Teoria del funcional de la densitat, DFT) per tal de modelar els metal·loful·lerens més interessants apareguts entre el 1999 i el 2005. Tot i que els metal·loful·lerens presenten una varietat d'estructures formidable, podem classificar-los en tres grans famílies des del punt de vista estructural: un primer grup són els metal·loful·lerens exoèdrics on els metalls es situen fora de la caixa, en un segon grup tenim els metal·loful·lerens heteroèdrics o heteroful·lerens on les caixes de carbonis han estat dopades (certes posicions han estat substituïdes) amb altres elements com ara: N, B, Si, Fe i altres metalls i finalment l'últim grup són els metal·loful·lerens endoèdrics on la derivatització metàl·lica es dóna a l'interior de la caixa. Així l'estudi teòric dels compostos més novedosos i novells de cadascuna de les tres famílies de metal·loful·lerens és el puntal de la present tesi doctoral. Per cadascuna de les famílies hem desgranat l'estructura electrònica i geomètrica, hem descrit el tipus d'enllaç metall-ful·lerè, hem avaluat els factors que intervenen en l'estabilitat relativa dels possibles isòmers, hem també predit la reactivitat davant de reaccions d'addició i finalment els hem caracteritzat des d'un punt de vista físic (càlcul del potencial d'ionització i afinitat electrònica). L'estructura electrònica ens ha permès seleccionar quins metal·loful·lerens endoèdrics seran estables avançant-nos als experimentalistes en la seva cerca de nous complexos. La diversitat de tipologies d'enllaç ha sigut tant gran com famílies de metal·loful·lerens. Hem descrit un enllaç iònic pels metal·loful·lerens endoèdrics, un enllaç covalent en el cas dels metal·loful·lerens heteroèdrics i un enllaç coordinatiu pels metal·loful·lerens exoèdrics. Hem aplicat noves metodologies per l'estudi de l'isomerisme. Les tècniques d'anàlisi multivariant de dades ens han permès esbrinar quins factors són importants per l'estabilitat relativa dels isòmers i a la vegada construir models de predicció per altres isòmers. Així doncs, l'acoblament de la química teòrica i la quimiometria ha estat sens dubte l'aportació més rellevant del present treball d'investigació.Tarragona, 28 de novembre de 2004Josep Maria Campanera AlsinaFINAL REPORT The characterisation of the most novel metallofullerenes up to 2004 has been theoretically and systematically discussed in this study. From the structural point of view, metallofullerenes can be divided into three main groups, all of which have been discussed in this study: endohedral, heterohedral and exohedral metallofullerenes. The main families of compounds studied are Sc3-nMnN@Ck (n = 0-3, M = Y, La; k = 68, 78, 80) (endohedral), CxMn (x = 56, 57, 58, 59; M = Pt, Ir, Os; n = 1, 2) (heterohedral) and (-Ck){M(PH3)2}n (k = 60, 70, 84; M = Pt, Pd, Ni; n = 1, 2, 4, 6) (exohedral). The present study is a step forward in our knowledge of each of these families of compounds, and in particular, in our understanding of the metal-carbon bond, isomerism and reactivity. The DFT method proved to be an excellent computational tool for providing good geometries, for solving the intricacies of the different metal-carbon bonds, for producing experimental data (ionization potentials and electron affinities) and also for making predictions about isomerism stability and reactivity. The principal conclusions drawn about the species studied here are: Different types of structures, different types of metal-carbon bonds. The metal units in each family of compounds are located differently in relation to the fullerene carbon framework: inside the cage, within the carbon framework and outside the cage. The encapsulation of a trimetallic nitride template unit (TNT, Sc3-nMnN; n = 0-3; M = Y, La) inside the carbon cage to produce TNT endohedral metallofullerenes is explained by an ionic pair (cage-metal) model in which the TNT unit formally transfers six electrons to the cage. On the other hand, in heterohedral metallofullerenes, metals establish a covalent metal-carbon bond without causing oxidation to the metal. Finally, the (MPH3)2 metal units situated exohedrally to the fullerene are only coordinated in a  mode to the CC bond. Chemometric tools applied to isomerism studies. The regioisomers of heterohedral metallofullerenes are numerous: for example, the stoichiometry C57Pt2 has 47 distinct regioisomers. Thus, chemometric techniques which can manage considerable amounts of data must be used if we want to understand regioisomerism in heterohedral metallofullerenes. These tools have also been very useful for drawing conclusions from the considerable quantities of data provided by the factors which affect the stability of regioisomers. These tools have been used not only for analysing data but also for predicting the stability of other heterofullerenes.TNT encapsulation stabilizes fullerene isomers that are not available as free fullerenes. TNT endohedral metallofullerenes are formed by the encapsulation of a metallic nitride template inside the following cages: D3-C68:6140, D3h'-C78:5, D5h-C80:6 and Ih-C80:7. Any of these cages have never been detected experimentally.So, endohedral metallofullerenes can make non-classical fullerene isomers available for study. Furthermore, on the basis of the electronic structure we predicted that no other IPR fullerenes between C60 and C84 will be capable of encapsulating a TNT unit, apart from the fullerene isomers that are already known.Stability of the carbon skeleton is the principal factor that determines the regioisomer stability of the heterofullerenes. Metal atoms occupy neighbouring positions in the most stable structures of C57Pt2 and C56Pt2. Metal substitution deforms the carbon framework and partially destroys the fullerene aromaticity. This is the key factor in determining the stability of these disubstituted clusters. Indeed, it is much easier to make a big hole that permits the incorporation of two Pt atoms in the carbon cage than two smaller holes in two opposite sites of the fullerene. Prediction of the exohedral reactivity taking into account the full characterization of the different CC bond types. We first performed a full characterization of all CC bond types of the fullerenes Ih-C60:1, D3-C68:6140, D5h-C70:1, D3h'-C78:5, Ih-C80:7, D2-C84:22 and D2d-C84:23. Each CC bond type is characterized by its topology, length, pyramidalization angle and Mayer bond order. This systematization enabled us to identify which sites were most reactive to a nucleophilic addition to free fullerenes or a [4 + 2] cycloaddition to TNT endohedral metallofullerenes.Tarragona, 8th Novembre 2004Josep Maria Campanera Alsina

This dissertation investigates three topics in the field of multiple-bonded metal chemistry. The first topic concerns the synthetic and theoretical considerations of ditungsten formamidinates and guanidinates compounds. This work presents an enhanced synthetic path to the W2(hpp)4 molecule (Hhpp = 1,3,4,6,7,8-hexahydro-2H-pyrimido[1,2-a]pyrimidine). The reflux of W(CO)6 with Hhpp in o-dichlorobenzene at 200 oC produces W2(hpp)4Cl2 in a one-pot reaction in 92% yield. This compound is stable and easily stored for further use, and it can be efficiently reduced in a one-step reaction to the most easily ionized compound W2(hpp)4. This work also examines the electronic structure and geometry of the intermediates W2(µ-CO)2(µ- hpp)2(η2-hpp)2 and W2(hpp)4Cl2. The second topic concerns the theoretical investigation by DFT of the electronic structure of [Mo2] units bridged by oxamidate ligands or bridging hydride ions ([Mo2] = (Mo2(DArF)3, where DArF is the anion of a diarylformamidine). It is shown that the effect of the gauche conformation of the α oxamidate isomers is due to steric interactions, and that the planar β oxamidate isomers have an electronic structure similar to that of naphthalene when it is doubly oxidized. The [Mo2](µ-H)2[Mo2] compound shows interdimetal unit interactions between the δ orbitals of the two [Mo2] units. These interactions are theoretically predicted and experimentally observed by a decrease in the [Mo2]---[Mo2] distance with a one-electron oxidation of [Mo2](µ- H)2[Mo2]. The final topic concerns the magnetic and structural properties of two Ru2(DArF)4Cl compounds. The compounds with Ar = p-anisyl (para) and m-anisyl (meta) both show different temperature dependence of their molar magnetic susceptibility, χ. For the para compound, there is a Boltzmann distribution between a π*3 ground state and a δ*2π* upper state, and this is confirmed by a temperature dependence of the Ru-Ru bond length: 2.4471(5) àat 23 K and 2.3968(5) àat 300 K. For the meta compound, a δ*2π*configuration persists over the range of 23-300 K as shown by an invariant Ru-Ru bond length and its molar magnetic susceptibility.

Chiral phosphoric acids are bifunctional catalysts that have the ability to activate electrophiles and nucleophiles through hydrogen bonding, and they have been successful in catalysing highly enantioselective additions of a wide range of nucleophiles to imines. In most literature reports it is not frequently revealed how these catalysts impart enantioselectivity. Thus, the vast majority of time required for reaction development is expended on the optimisation of the catalyst features. The research described here explores the ability of relating computational derived catalyst parameters to enantioselectivity as a means to assess the catalyst features important for enantioinduction. The proposed features are evaluated computationally and summarised into simple qualitative models to understand and predict outcomes of similar reactions. In Chapter 1, I provide an overview of the progress and challenges in the development of chiral phosphoric acid mediated reactions. I highlight leading computational studies that have enabled a greater understanding of how the catalyst imparts reactivity and selectivity. In general, the studies focus on the most effective catalyst and do not do a detailed investigation into the effects of changing the substituents at the 3,3’ positions. Implicating steric effects from reasonably large groups as a key component in imparting enantioselectivity. However, it is clear that they have a more subtle effect. A large group is required but if it is too large poor or unusual results are obtained, making the correct choice of reaction conditions challenging. In Chapter 2, I develop a quantitative assessment of the substituents at the 3,3’ positions. I show in Chapters 3 and 4 that I can use rotation barriers in combination with a novel steric parameter, AREA(θ), to correlate enantioselectivity. By exploiting this finding, the catalyst features important for enantioselectivity can be identified, and this is validated by QM/MM hybrid calculations. Summarising these detailed calculations into a single qualitative model, guides optimal catalyst choice for all seventy-seven literature reactions reporting over 1000 transformations. These mechanistic studies have guided the design of a new catalyst with increased versatility, which is discussed in Chapter 5. Chapter 6 details my study into the effect of the hydroxyl group on the mechanism of transfer hydrogenation of imines derived from ortho-hydroxyacetophenone. I show, using detailed DFT and ONIOM calculations, that transition states of these reactions involve hydrogen bonding from both the hydroxyl group on the imine and the nucleophile’s proton to the phosphate catalyst. In Chapter 7, computational analysis is used to provide insight into the origins of enantioselectivity in chiral phosphoric acid catalysed Friedel-Crafts and Mannich reactions proceeding through monoactivation mechanisms. The final chapter contains an in-depth look into the stereoelectronic effects altering enantioselectivity in the silver-phosphate mediated spirocyclisation reaction involving aromatic ynones. In this study I show that enantioselectivity is governed by the non-covalent interactions between the aromatic group of the ynone and the 3,3’ substituent. I was able to propose synthetic modifications to the substrate used in this reaction, resulting in an improvement in enantioselectivity.

No description available.

NMR spectroscopy with in situ laser irradiation has been used to investigate the photo- and thermal isomerisation of eight azobenzene derivatives; diphenyldiazene (azobenzene), p-phenylazoaniline (p-aminoazobenzene), 4-(dimethylamino)azobenzene (Methyl Yellow), 4-dimethylamino-2-methylazobenzene (o-Methyl-Methyl Yellow), p-nitroazobenzene, 4-nitro-4’-dimethylaminoazobeneze (Dimethyl-nitroazobenzene), 4-(4-nitrophenylazo)aniline (Disperse Orange 3) and N-ethyl-N-(2-hydroxyethyl)-4-(4-nitrophenylazo) (Disperse Red 1). The rate constants and activation parameters of the thermal cis-to-trans isomerisation have been measured experimentally and correlated to the mechanism of isomerisation in two solvents. The experimental data show that the values of the activation energy (related to the enthalpy of activation) and the entropy of activation (related to the Arrhenius pre-exponential factor) vary significantly from molecule to molecule and thus both of these parameters influence the inter-molecule variation of the rate constant. Similarly, both of these parameters influence the solvent-dependence of the rate constant. Complementary computational studies have been carried out in the gas phase and in solution using density functional theory (DFT) to predict the structures of the cis and trans isomers and the transition state, and to explore the reaction coordinate. The theoretically predicted activation parameters are compared with those determined experimentally, and the utility of DFT calculations in predicting the effects of molecular structure and solvation on the kinetics of cis-to-trans isomerisation assessed. The DFT-predicted values of the activation energy and Gibbs free energy of activation in DMSO are in good agreement with the experimental values, while the values in benzene tend to be in less good agreement. The DFT calculations are unsuccessful at predicting the entropy of activation, where in all cases there is a large discrepancy between the theoretical and experimental values. The DFT- calculated energy differences between the activation energies of the two inversion pathways for the asymmetric azobenzenes suggests the favourable phenyl ring for inversion. The formation of a linear transition state from a dihedral rotation potential energy curve is explained in terms of the lower activation barrier of the more favourable inversion route (α-inversion) than that of the dihedral rotation pathway, and suggests the inversion through the α-phenyl ring to be the favoured pathway for substituted azobenzene. DFT calculations are able to obtain a transition state corresponding to pure rotation pathway for two azobenzene derivatives. The higher activation barrier for the formation of the transition state corresponding to this pathway compared to that of the formation of the α-transition state confirmed the previous conclusion. DFT predictions of the effect of protonation on the thermal rates of isomerisation of azobenzenes substituted with electron-donating group were in good agreement with the experimental results; both conclude faster isomerisation and lower activation barriers on protonation. However, DFT calculations could not confirm the postulation of rotational transition state for the isomerisation of the protonated molecule, as a result of weakening of the N=N bond by protonation.

Clusters constitute an intermediate state of matter between molecules and solids whose properties are size dependent and can be tailored. In recent years, cluster science has become one of the most exciting areas of research since their study can not only bridge our understanding between atoms and their bulk but also between various disciplines. In addition, clusters can serve as a source of new materials with uncommon properties. This dissertation deals with an in-depth study of clusters as a bridge across physics, chemistry, and materials science and provides a fundamental understanding of the structure-property relationships by focusing on three different topics. The first topic deals with superatoms which are clusters that mimic the chemistry of atoms. I show that superhalogens and superalkalis can be designed to mimic the chemistry of halogen and alkali atoms, respectively. An entirely new class of salts can then be synthesized by using these superatoms as the building blocks. I have also explored the possibility of designing highly electronegative species called hyperhalogens by using superhalogens as ligands or superalkalis as core and a combination of both. Another aspect of my work on superatom is to examine if traditional catalysts (namely Pd) can be replaced by clusters composed of earthabundant elements (namely Zr and O). This is accomplished by comparing the electronic structure and reactivity of Pd clusters with isoelectronic ZrO clusters. The second topic deals with a study of the electronic structure of coinage metal (Cu and Ag) clusters and see if they remain unchanged when a metal atom is replaced by an isoelectronic hydrogen atom as is the case with Au-H clusters. The third topic deals with clusters as model of polymeric materials to understand their gas storage and sequestration properties. This is accomplished by studying the trapping of H2, CO2, CH4 and SO2 molecules in borazine-linked polymers (BLPs) and benzimidazole-linked polymers (BILPs). The first two topics provide a bridge between physics and chemistry, while the third topic provides a bridge to materials science.

This thesis investigates the stabilization of divalent germanium using substituted diethanol amine ligands. Germylenes of type RN(CH2CH2OH)2Ge were obtained from N-heterocyclic germylenes and N-alkyl diethanol amines in yields of up to 94%. Single crystal X-ray diffraction confims the presence of a transannular Ge-N dative bond in all cases. In addition, intermolecular dimers containing Ge2O2 rings are formed for R = Me and Et. Reaction of the four germylenes L with nickel carbonyl yielded the respective germylene complexes L2Ni(CO)2 and LNi(CO)3. The germylenes and their complexes were investigated with DFT methods. Only four methods, SVWN, BB1K, MPWB1K and M062x gave acceptable Ge-N distances. Dimerization energies of the germylenes were examined with the thermochemically accurate M062x method. At the M062x/Def2-TZVP level, the dimerization energies of the germylenes are very small (ΔG° ≈ 0 kcal/mol). The experimentally observed dimerization or lack thereof may accordingly be determined by packing effects in the solid state or solvation energies in solution.

The vibrational entropy calculated by applying the harmonic oscillator approximation to all vibrational degrees of freedom is inherently inaccurate. One major reason is because low frequency modes such as internal rotations are not properly described by this approximation. Various techniques were developed in the past to overcome this problem. The hindered rotor potential can be approximated by a series of cosine functions, and the relevant coe cients can be determined by tting to a calculated potential energy surface. However, such a method is di cult and time consuming. Therefore, in this dissertation we propose and describe two less tedious approaches to determine entropy of internal rotational modes. The rst approximation is to express the barrier height in terms of the harmonic oscillator frequency, the local periodicity, and the reduced moment of inertia of the rotation and to approximate the torsional potential by a single cosine function. Thus, the 1D Schr odinger equation for internal rotations can be solved without nding the torsional potential, transition states, or barrier heights. We propose a further simpli cation to this approach, achieved through a simple mathematical formula, that interpolates the hindered rotor entropy between the free rotor and harmonic oscillator limits. We also propose a procedure to automatically determine the axis of rotation for any hindered rotor. The proposed methods were applied to determine the torsional entropy of n- alkanes from ethane to hexane. The entropies calculated from the proposed methods give good agreement with the experimental and accurately calculated values and have a signi cantly better accuracy than the harmonic oscillator approximation. Furthermore, we performed approximate and full hindered rotor treatments to nd the corrected vibrational entropy of bis(chromiumtricarbonyl) dibenzo[a,e]cyclooctatetraene (DBCOT). The eighth chapter of this dissertation is an independent molecular dynamics (MD) project to study how ethanol interacts with human and mouse Toll-Like- Receptor3 (TLR3) monomers and a TLR3-dsRNA complex. No major structural changes were observed during the ethanol docking and subsequent MD simulations, but the MD simulations revealed a reduction in the proportion of alpha helix present during a 1000 ns MD simulation on the h-TLR3 monomer in 0.5 percent ethanol.

<p>With increasing global renewable energy demands, there is a need for new materials with improved performance, lower cost, and less toxicity. One such application is photovoltaics, where halide perovskites (HaPs) represent the fastest growing market of absorbers owing to their impressive optoelectronic properties and excellent tunability from composition engineering and structural manipulation. However, the practically infinite composition-structure space of HaPs when considering cation and/or anion site mixing, octahedral distortion and rotation, and other forms of polymorphism, raise considerable challenges when comprehensively exploring their stability and optoelectronic properties. First principles calculations are powerful tools that can investigate large numbers of compounds and structures in a high-throughput fashion. </p><p>In my thesis, I performed high-throughput density functional theory (DFT) computations to generate a HaP dataset within a wide chemical space covering ~500 unique chemical compositions in the (pseudo-)cubic phase, across a 14-dimensional ionic space. This work explored both pure and alloyed compositions, with the latter simulated using the special quasi-random structures approach. Many critical properties were computed using the semi-local GGA-PBE and hybrid non-local HSE06 functionals, including decomposition and mixing energies, electronic band gap, and spectroscopic limited maximum efficiency (SLME), which is a theoretical surrogate for the likely absorption efficiency of the compound when used in a single-junction solar cell. Property screening over this dataset yielded 32 stable perovskite candidates with attractive optoelectronic properties.</p><p>Polymorphism in HaPs is investigated by simulating larger supercell alloys with different ionic ordering, generating compounds with random octahedral distortions and rotations, and optimizing various compositions in non-cubic phases such as tetragonal and orthorhombic. Linear correlation analysis is performed to gain a critical understanding of how properties are influenced by specific cations and anions, their mixing fractions, the perovskite phase, ionic clustering, and amount of strain or distortion in the lattice. Finally, trends, design rules, and predictive insights achieved from the DFT datasets are applied over a much larger set of thousands of hypothetical compounds, resulting in identification of more promising materials and understanding of the most important A-B-X combinations that yield multiple desired objectives.</p><p>Furthermore, a similar DFT workflow is applied for designing transition metal electrocatalysts. DFT simulations are performed to model Hydrogen adsorption, OH adsorption, and the water splitting reaction on Ni3N/Ni and Co2N/Co hybrid structures, to explore their likelihood in being used for Hydrogen Evolution Reaction (HER). The results reveal the excellent catalytic performance of transition metal and transition metal nitride hybrid structures.</p><p><br></p>

No description available.