No abstract

A molecular beam of phenol, cooled by a supersonic expansion, is crossed at right angles by the output of a pulsed frequency-doubled dye laser, causing 1 + 1 resonance enhanced multiphoton ionization. The kinetic energy of the resulting photoelectrons is determined as a function oflaser wavelength with time-of-flight analysis, permitting the assignment of 11 vibrational frequencies for the 2BI phenol-h 6 cation and ten vibrational frequencies for phenol-d 5. Of these, all but the lowest frequency one in each case are in-plane vibrations of which phenol has a total of 19. An approximate harmonic force field for the in-plane modes of the phenol cation is derived along with its associated frequencies and mode forms. This in turn facilitates the vibrational analysis. Analogous force field calculations have been carried out on the ground eA d and first excited (I B 2) states of the neutral parent, permitting conclusions to be reached concerning bonding changes upon removal of an electron from the phenol electron system.

The accuracy of vapor phase vibrational data has been improved for all 12 deuterium-labeled benzenes and for 13C12C5H6 and 13C6H6. Many vapor phase fundamental frequencies are observed for the first time. Precise isotopic frequency/splitting patterns for ν1, ν18, and ν19 have been obtained. Isotope induced harmonic mode mixing matrices are given for all 14 labeled benzenes and used to provide detailed description of the fundamental bands observed in the spectra. These descriptions provide numerous reassignments for the fundamental bands, particularily in low symmetry deuterium benzenes. The matrices show that some skeletal modes, such as ν1, gain CH stretching character as a result of deuterium labeling, providing a rationalization for the increased anharmonicity observed in recent jet experiments for C6D6. In addition, a reassessment of Fermi resonance gives 3072.3 cm−1 for the unperturbed frequency (correction +24 cm−1) for the e1u mode ν20 in C6H6 refining the CH local mode anharmonic constant, 2xii, to 117.5 cm−1.

Fully-relaxed model ab initio calculations at Hartree–Fock/6–31G(d,p) and Mo/ller–Plesset (MP2)/6–31G(d,p) levels for acetaldehyde methyl conformers indicate significant skeletal flexing (e.g., the CH3 –C bond length changes by 0.006 Å) and methyl hydrogen folding. Thirteen methyl conformer energies at 15° intervals are used to assess the magnitudes of the torsional potential function expansion terms. Only two terms V3=373.8 and V6=3.4 cm−1 (both significantly different from those obtained from microwave and infrared analyses) are found to be important. These calculations clearly show that relaxation during methyl rotation (i.e., skeletal flexing and methyl hydrogen folding) is an important determinant of the torsional potential. Energy levels obtained from internal rotation potentials which include flexing simulate infrared torsional fundamental frequencies in CH3CHO and CD3CHO to within 1–2 cm−1 of the experimental values. In the absence of relaxation infrared torsional fundamental frequencies are poorly simulated by ab initio calculations.

The ground-state potential surface for methyl torsional interactions in acetone has been obtained by ab initio, semiempirical, and empirical methods. Frequencies acquired from recent Rydberg spectroscopy measurements on the a2 ground-state fundamental and overtone torsional vibrations and infrared measurements on b2 fundamentals in acetone-A6 and acetone-d6 are well simulated by purely theoretical ab initio HF/6-31G(d,p) and scaled MP2/6-31G(d,p) (i.e., semiempirical) fully-relaxed model potentials. This model incorporates skeletal flexing during the methyl rotation, causing the CCC angle to vary by 3°. The single unique empirical surface obtained from the above-mentioned measured frequencies is in good agreement with the ab initio and scaled ones. The four empirically determined constants in the two (equivalent C3") top potential function, eq 3, are V3 = 370 cm"1, V33 = 136 cm"1, V'33 = -156 cm"1 and V6 = 0 cm"1. The torsional barrier, Keff, is found to be 240 cm'1, significantly lower than the microwave value. We are unable to predict potential constants that simulate the frequency-generated empirical ones at any level of ab initio calculation [i.e., HF/DZ, HF/6-31G(d,p), and MP2/6-31G(d,p)] within the rigid-frame model (i.e., geometry parameters at equilibrium are held fixed during the methyl top rotation).

No abstract

We demonstrate that fundamental frequencies provide a poor criterion of the benzene B2u force field accuracy and that two-photon cross sections of the b2u fundamental bands in the 1B2u↔1A1g electronic transition, which can be directly related to skeletal displacement magnitudes in the two b2u modes, provide an insightful physical criterion of harmonic force field quality. Another valid criterion for force field quality is isotopic frequency shifts combined with the fundamental frequencies. The frequency-generated force field of part II accurately predicts the measured cross sections and isotopic frequency shifts, indicating that the B2u force constants are known to ±0.01 mdyn/Å. These constants are used as benchmark quantities for calibrating theoretically modeled force fields. A systematic series of ab initio B2u harmonic force fields for ground state benzene using theoretical geometries are generated at Hartree–Fock and correlated second, third, and fourth order (with single, double, triple, and quadruple excitation) Mo/ller–Plesset perturbation theory (MP2, MP3, MP4SDTQ) and configuration interaction theory with all single and double excitation (CISD) levels using basis sets from minimal double zeta to triple zeta plus diffuse and polarized functions. These theoretical models of the B2u force field all provide poor predictions for the three criteria: fundamental frequency accuracy 2%–3%; isotopic frequency shift accuracy 10%–300%; two-photon cross section accuracy 300%–1200% with the sense of isotopic effects on two-photon cross sections in some cases incorrectly predicted. The MP2 calculations, even using the largest basis set, are incapable of meeting any of the criteria, hence higher order approaches to the correlation problem are required. The inadequacies in frequencies, isotopic shifts, and mode forms arise because both the diagonal and off-diagonal force constants are not predicted by ab initio calculations with the sufficient 10−2 mdyn/Å accuracy required for reasonably accurate frequency and intensity predictions. A feature of the ab initio calculations is that carbon and hydrogen displacement phases for the b2u modes are unchanged by the basis set size or correlation level. The unmeasured 13C6D6ν14 two-photon cross section and iosotope frequency shift from 12C6H6 are predicted to be larger than for any of the other D6h symmetry benzenes (∼30% higher than in C6H6 for the former and 72 cm−1 for the latter) by the benchmark field of part II.

No abstract