Rotationally resolved spectra of gas-phase samples of 1-(1-naphthyl)-ethylamine (NEA) and amine deuterated forms have been obtained in the microwave and ultraviolet regions, with the isotopomers initially prepared in their zero-point vibrational levels by cooling in pulsed jet and molecular beam supersonic expansions. A single parameter set that includes inertial parameters and 14N nuclear quadrupole constants has accounted for nearly all transitions observed in the Fourier-transform microwave spectrum at 2 K and indicates the presence of only one geometrical isomer in the jet-cooled expansion. The rotational constants, dipole moment orientation, and the amine hydrogen atom positions have been used to identify the conformation of the attached chiral group from among nine possible isomeric forms. The S 1 rotational constants and electronic transition moment orientation have been obtained from high resolution molecular beam data of the band origin at 31771.56(2) cm −1. Excited state “gas phase” predictions from ab initio theory at the CIS/6-31G(d,p) and CIS/6-311 + G(d) levels are compared with the observed S 1 results and with circular dichroism (CD) data obtained for solution phase samples of ( S)-NEA in cyclohexane. Distinguishing spectral features are found in the calculated CD spectra of the three lowest energy conformers arising from rotation about the C–CH(NH 2)CH 3 bond, all of which are thermally populated at room temperature. While the predictions at both levels are in fair agreement with the observed gas phase results, CIS/6-31G(d,p) theory is found to be inadequate to model the condensed phase CD spectrum. In constrast, the calculated spectrum of the equilibrium mixture of conformers at the CIS/6-311 + G(d) level is in good qualitative agreement with the observed CD results and indicates the importance of diffuse functions for the accurate prediction of chiroptical properties of NEA. The most noteworthy exception is the predicted rotatory strength of the S 1 origin, which is overestimated by more than 10-fold. Possible reasons for this discrepancy with experiment include a reversal of S 1(L b) and S 2(L a) states from the absence of electron correlation and/or the neglect of room-temperature solvent effects. These results provide for rigorous benchmark tests of theoretical models and further elucidate the importance of the conformational structure for determinations of absolute stereochemistry from CD spectra.
Read full abstract