|
|
[2.2]Paracylophane:
A Vibrational Study of a Strained Organic Molecule Using
Inelastic Neutron Scattering Spectroscopy
|
[2.2]Paracyclophane is the parent compound in the family of cyclophanes, see Figure 1. This molecule consists of two benzene rings on top of each other in parallel planes and connected by ethano bridges at the para positions. The bridge hydrogens create strain in the molecular structure, making it an interesting case study for vibrational spectroscopy. A large number of low frequency vibrations and limited IR and Raman data qualify inelastic neutron scattering spectroscopy as an excellent tool for studying this molecule. Past vibrational studies have produced varied and uncertain interpretation of their results[1-4]. |
|
||
|
Geometries of Paracyclophane Early crystallographic studies found the crystal structure of [2.2]paracyclophane to have D2h symmetry [5]. Semi-empirical calculations further confirmed these findings[6-10]. Through reexamination of the crystallographic data and new quantum mechanical calculations the geometry of the molecule was later found to have D2 symmetry. In the D2 symmetry, the bridge hydrogens are staggered, not eclipsed, and the benzene rings are slightly twisted with respect to each other about an axis perpendicular to their planes, see Figure 1. The molecular geometry has important consequences for the vibrational properties of the molecule. [2.2]Paracyclophane has ninety normal modes of vibrations. For D2 symmetry all ninety vibrations are seen in the Raman spectrum and sixty-six (B1, B2, and B3) vibrations are IR active. Since the molecule deviates only slightly from the D2, the IR/Raman peaks owing their intensity to this deviation are very weak. The combination of numerous active modes and the fact that many of them are very weak hinders a secure vibrational assignment using IR/Raman data alone. The twist which changes the molecular symmetry from For D2 to D2 can can occur in both the clockwise and counter-clockwise direction. This result in two equivalent minimum energy geometries. The harmonic approximation used in current computations does not take this double energy minima into effect. This uncertainty in the normal mode calculations adds to the problems in making confident a vibrational assignment. |
|
||
|
Inelastic Neutron Scattering Spectroscopy Unlike IR or Raman spectroscopy, INS spectroscopy has no selection rules. This makes it possible to observe vibrations previously never seen, providing more data for a vibrational study. A major advantage of INS spectroscopy in comparison to IR and Raman spectroscopy is that a reliable, easily implemented theory exists for computation of INS intensities. The focus of this paper is primarily on the lower frequency range of the spectrum. In the higher frequency region a quantitative treatment must also include combinations and overtone transitions. The INS experiments on [2.2]paracyclophane were completed utilizing the Filter Analyzer Neutron Spectrometer (FANS) at the NIST Center for Neutron Research (NCNR) [11]. The experiments were done at a temperature of 10 K and over two overlapping frequency ranges, see Figure 3. The spectrum from 40 wavenumbers to 325 wavenumbers was taken with a polycrystalline graphite monochromator, and the spectrum from 200 wavenumbers to 1200 wavenumbers was taken using a copper monochromator. The INS spectrum is compared with both IR and Raman data in Figure 4. The vibrations seen in the IR/Raman spectra are also seen in the INS spectrum, however many vibrations unseen in the other spectra are clearly seen in the INS spectrum. It should be noted that the peaks below 100 wavenumbers are due to spectrometer artifacts and are not to be considered real intensity. These are a consequence of the FANS instrument and future experiments utilizing a different machine will eliminate this problem. The correction is important since calculations show that two real peaks should be present in this frequency range, see Figure 5. |
|
||
|
Theoretical Comparison The calculated data shown in Figure 5 is based on a restricted Hartree-Fock geometry optimization and normal mode calculation using the STO-3G basis. The calculated normal mode frequencies are shown as small peaks with narrow bandwidths for visualization purposes but with relative amplitudes that reflect the calculated intensities. As Figure 4 makes evident, the current calculation fails to produce both accurate frequencies and intensities. The current calculation uses the small basis set, STO-3G, which frequently produces inaccurate results. However, the error in the current calculation is greater then usually seen for the STO-3G basis, indicating other reasons contributing to the error. A likely source of this additional inaccuracy lies in the possible expected anharmonicity of the low frequency vibrations not accounted for in the present calculations. |
|
||
|
Future Investigations and Experiments Future investigations will include calculations using improved methods including those drawing on density functional theory and Møller-Plesset correlation energy corrections, all of which will be done with a variety of basis sets. A periodic lattice calculation will also be completed which will account for any intermolecular interactions in the crystal. This calculation uses crystal structure data to perform a normal mode calculation equivalent to performing a calculation on the whole crystal. A future vibrational experiment is planned using a time-of-flight neutron scattering spectrometer. This will eliminate the intensity to quasielastical scattering as well as provide better resolution at low frequency. A neutron diffraction experiment may be possible if a fully-deuterated [2.2]paracyclophane sample can be obtained. |
![[2.2]Paracyclophane, C16H16](images/paracyclophane1.jpg)
|