构象异构
化学
基体隔离
二面角
密度泛函理论
异构化
红外光谱学
结晶学
计算化学
氢键
分子
有机化学
催化作用
作者
Ana Borba,Andrea Gómez‐Zavaglia,Pedro Simões,Rui Fausto
摘要
The preferred conformations of dimethyl sulfite and their vibrational spectra were studied by matrix-isolation Fourier transform infrared spectroscopy and theoretical methods (density functional theory (DFT) and Moller-Plesset (MP2), with basis sets of different sizes, including the quadruple-zeta, aug-cc-pVQZ basis). Five minima were found at these levels of theory. At the MP2/6-31++G(d,p) and DFT/B3LYP/aug-cc-pVQPZ levels, the GG conformer (where the O-S-O-C dihedral angles are 73.2 and 70.8 degrees ) resulted in the conformational ground state. At the highest level of theory used, the GT conformer (O-S-O-C = +68.5 and -173.2 degrees ) is 0.83 kJ mol(-1) higher in energy than the GG form, while conformer GG' (O-S-O-C = +85.7 and -85.7 degrees ) has a relative energy of 1.18 kJ mol(-1). The remaining two conformers (G'T and TT) are high-energy forms and not experimentally relevant. In consonance with the theoretical predictions, conformer GG was found to be the most stable conformer in the gaseous phase as well as in the low-temperature matrices. Annealing of the argon matrices first promotes the GG'-->GT isomerization, which is followed by conversion of GT into the most stable conformer. There is no evidence of occurrence of GG'-->GG direct conversion in the low-temperature matrices. On the other hand, during deposition of the xenon matrices conformer GG' totally converts to conformer GT. Two observations demonstrated this fact: no evidence of bands corresponding to GG' were observed in xenon matrices and the GG/GT intensity ratio became similar to the GG/(GT + GG') intensity ratio observed in argon matrices. All these results could be explained by taking into account the relative values of the theoretically predicted energy barriers for the different isomerization processes: GG'-->GT, 1.90 kJ mol(-1); GT-->GG, 9.64 kJ mol(-1); and GG'-->GG, 19.46 kJ mol(-1).
科研通智能强力驱动
Strongly Powered by AbleSci AI