Hydrogen abstraction reactions [Formula: see text] with R, X≡H, CH3, NH2,OH, and F have been studied at the abinitio 6-31G – UHF level. However, energetic properties were computed at the CI level. Rate constants and Arrhenius parameters have been obtained using the transition state theory formalism with Eckart's tunneling correction. The discrepancy between theoretical and experimental results led us to elaborate a semi-empirical procedure to calculate activation barriers, in which the bonds R—H and X—H are represented by Morse curves. Thus, the agreement between theory and experiment is much better. Moreover, the results obtained by this procedure demonstrate the non-Arrhenius behavior of all the reactions under consideration and allow us to rationalize a large number of experimental facts.

For a set of small XYH, singlet-state molecules (where X stands for Li, Be, or B and Y is one of the second-row atoms), we have calculated the enthalpies of formation at the MP4(sdtq)/6-31l++G(3df,Zp) level by using MPP(full)/6-31G(d,p) fully optimized structures. At this level of theory, the heats of formation are expected to be in the range of the so-called "chemical accuracy" (fl kcal/mol). Some alkyl derivatives of the previous XYH, compounds have been studied at the RHF/6-31G(d) level (with fully optimized geometries). The theoretical enthalpies of formation reproduce the available experimental results quite satisfactorily. In addition to compounds with usual dative single bonds, we also describe molecules with covalent (nondative) single B-N and B-0 bonds. We use all the available (experimental and theoretical) data to calculate new bond energies for lithium, beryllium, and boron derivatives. IntroductionAlthough many features of electron-deficient compounds have been established by a number of experimental and theoretical investigations, accurate thermochemical properties remain undetermined, at least for the simplest derivatives. Except for LiOH, BeHF, and BH20H, very few enthalpies are known for lithium-, beryllium-, and boronbased materials (XYHJ. Most often, those species cannot be handled easily. For example, BH3*NH3 is stable but not easily vaporized2 and it rapidly decomposes when the temperature increases. BH2NH2 is un~table;~ its lifetime was reported to be several minutes, and it prefers to exist aa cyclic oligomers or polymers: BHNH may have a fairly long lifetime under arc discharge5 but is difficult to observe under other conditions. Therefore, direct determination of the energy content for such species is not easy. Nevertheless, a few alkyl and fluoro derivatives of the previously mentioned XYH, molecules have been investigated with success.From a theoretical point of view, difficulties arise because correlation effects account for a significant part of the energy content and can become a dominant part of reaction heats? Therefore, it is not surprising to observe large discrepancies in the calculated dissociation energies depending on the theoretical method in use. This has specially been pointed out for dib0rane(6)~J and for molecules with weak dative bonds. For example, the dissociation energy of ammonia-borane is predicted to be 46 kcal/mol at the HF/6-31G level, 21 kcal/mol at the HF/6-31G* level,s and 35 kcal/mol when the correlation energy is taken into a c~o u n t .~The aim of this work is to calculate accurate ( f l kcal/mol) heats of formation for the smallest lithium-,

No abstract

No abstract

The potential energy hypersurface for the isomerization of azidoazomethine to 1//-tetrazole has been studied by ab initio calculations using an STO-3G basis set. The reaction path for this cyclization is such that the heavy atoms remain in the same plane. The formation of the activated complex results from a movement of the lone pair on the NH group toward the terminal nitrogen atom. At the same time, a lone pair is formed on the central nitrogen atom of the azide group at the expense of the ir NN bond and the system is scarcely altered. The transition state was therefore seen to resemble strongly the reactant azide rather than the tetrazole. A keynote in this process is the role of the lone pair on the NH group which permits the formation of the new bond. It can be predicted that polar solvents will preferentially stabilize the transition state which is more polarized with respect to the azide form. Thus, the cyclization should be accelerated in polar solvents. Our theoretical results are in reasonable agreement with available experimental ones.