A combined analysis of the nu(9) band and the far-infrared torsional spectra of ethane

Fourier transform measurements of the lowest frequency degenerate fundament al band of CH3CH3 (upsilon(9) = 1 <-- 0) in the 12-mu m region together wit h far-infrared torsional spectra have been analyzed to investigate vibratio n-torsion-rotation effects in a symmetric top molecule. Several spectra of the nu(9) band were recorded under different experimental conditions with a podized slit functions of about 0.002 cm(-1). Although the intrinsic tunnel ing splitting in each (J'<-- J') doublet in the nu(9) band is predicted to be of the order of 0.002 cm(-1), in some cases the observed splitting for a n intermediate J' of 20 is several times this value. In extreme cases, spli ttings of the order of 0.25 cm(-1) have been observed. These splittings are caused primarily by the Coriolis interaction between the torsional stack o f levels upsilon(4) = 0,1,2,..., for upsilon(9) = 1 and the corresponding s tack for the ground vibrational state. Because of a near-degeneracy between the upper level in the nu(9) band and its interacting partner (upsilon(9) = 0, upsilon(4) = 3), the (l = -1;K = 17,sigma = 0) torsion-rotation series is resonantly perturbed. For this case, perturbation-allowed upsilon(4) = 3 <-- 0 torsional transitions have been identified. Here sigma = 0, 1, 2, o r 3 labels the torsional sublevels. Measurements from the nu(9) and 3 nu(4) bands, frequencies from the far-infrared torsional spectra in the ground v ibrational state, and lower state combination differences from nu(9) + nu(4 ) - nu(4) band were fitted to within experimental uncertainty using a symme try adapted effective Hamiltonian which has been used for analyses of simil ar spectra in methyl silane and CH3CD3. Two Coriolis parameters were determ ined: the experimental value of <(zeta)over tilde>(z)(9) = 0.2610(12) is in good agreement with the calculated value of 0.25, whereas the experimental value of <(zeta)over tilde>(x)(4,9) = 0.2267(20) is about 3 times smaller than the calculated value of 0.60. The theoretical treatment presented here makes use of standard symmetric top formalism and the G(36)(dagger) double -group formalism. (C) 1999 American Institute of Physics. [S0021-9606(99)01 544-5].