This document is © Elsevier Science, 1994
Bg = h/(8p²·Ig), g = a, b, c
where Ig is the moment of inertia with respect to the principal inertial axis g. The moment of inertia is given by
Ig = S(mi·rgi²)
where mi signifies the mass of atom i and rgi the distance of the ith atom from the g axis.
Therefore, to predict the rotational spectrum of a certain molecule - which is an almost indispensable prerequisite for assigning the spectrum - without having relevant data available from other methods one needs to estimate the molecular geometry of the molecule.
In the course of our investigations in the field of cyclic cross-conjugated compounds predictions for rotational constants had to be made frequently. Since no useful experimental data were available for most of the molecules, we had to investigate the molecular geometries of these compounds theoretically.
There are two basic approaches for such an investigation. First, one can adapt parameters from similar molecules that have been determined experimentally assuming them to be the same in the molecule under study. As there is a considerable bandwidth for most parameters, the rotational constants obtained by this method are unlikely to be accurate enough in most cases.
With the advent of fast computers a second possibility has become feasible: the calculation of geometries by empirical forcefield or quantum mechanical methods. The accuracy of these geometries - and the resulting rotational constants - is the objective of this paper.
D = Ic - Ia - Ib
and should be close to zero for planar molecules. The results were planarity for both cyclopentadienone (Ia) [1,2] and the methylene analog fulvene (Ib) [3,4].
The situation with six-membered rings is more complex. The carbon atom in
4-position shows sp³-hybridization and so there are always at least
two atoms out of the ring plane. If the out-of-plane atoms are just two
hydrogens the defect of inertia can be evaluated in good approximation.
The corresponding molecules (IIa and IIb), however,
are not stable. We were unable to detect 2,5-cyclohexadienone (IIa) by microwave spectroscopy and the works on this molecule
that have been published so far [5,6] do not include
any structural information. The methylene analog 1-methylene-2,5-
cyclohexadiene1 (IIb) has been known for almost
three decades [7-9], but no structural information
was available until recently, when the molecule was shown to have a planar
ring by microwave spectroscopy [10].
A main reason for the instability of the two compounds
(IIa) and (IIb) is tautomerism. The carbonyl compound rearranges to
form phenol, while the methylene compound mainly forms toluene. One
possibility to prevent this rearrangement is to substitute both hydrogens
in 4-position by methyl groups. In the case of the ketone this leads to
4,4-dimethyl-2,5-cyclohexadienone (IIc), which was
also investigated by microwave spectroscopy [11].
The two methyl groups, however, make a decision about planarity impossible
on the basis of the rotational constants of one isotopomer alone. The
structure of the corresponding methylene compound (
IId) has been investigated by electron diffraction
[12] and the data were interpreted as indicative of a non-planar ring
with dihedral angles within the ring of about 8 degrees.
Theoretical calculations
We thought it would be worthwhile comparing the structural data we obtained
from the experiments with those predicted by theoretical calculations. Such
a comparison would provide information about the reliability of theoretical
results that can not yet be checked by experimental values. So we calculated
the structures of several cross-conjugated molecules with ab-initio methods
at two different levels of theory (RHF-6/31G* and STO-3G) using GAUSSIAN86
[13] and also semi-empirically with three different
Hamiltonians (MINDO/3 [14a]], MNDO [14b] and AM1 [14c]) using MOPAC [14d]. The calculations were carried out using the DEC
VAX 9410/6440 cluster of the computational center of the University
of Ulm. All the molecules considered in this study possess CS symmetry
if they are non-planar, whereas the symmetry is C2v if they are planar. So
we started with only CS symmetry imposed on the calculations.
2,5-Cyclohexadienone (cf. Table 1)
The initial parameters for cyclohexadienone (IIa) were taken from an optimized structure calculated by the molecular modeling package SYBYL [15a] using a TRIPOS-forcefield [15b]. The structure calculated by SYBYL was planar no matter if the starting-conformation was planar, boat, chair or half-chair. So we twisted the molecule arbitrarily to a conformation where the C6-C1-C2-plane and the C3-C4-C5-plane were at plane angles of 2 degrees and 10 degrees respectively to the C2-C3-C5-C6-plane. This conformation was then used as the initial conformation for the optimizations with only CS symmetry fixed.
All optimized geometries were planar within the convergence criteria and so we decided to restrain the molecule to the indicated C2v symmetry for a conformation that would be exactly planar without having to increase computing time overly.
4,4-Dimethyl-2,5-cyclohexadienone (cf. Table2)
For the dimethylated ketone (IIc) we chose the C2v optimized geometry of (IIa) with reasonable parameters for the methyl groups, but with only CS symmetry required. The resulting structures, again, were planar without doubt and so the same procedure as in the case of (IIa) was applied in order to obtain values for an exactly planar geometry.
Methylene compounds with six membered rings (cf. Tables 3 and 4)
In the case of the methylene-compounds (IIb and IId) the initial parameters used were those given in an electron diffraction study by Trætteberg et al. [12] for (IIb) (i.e. dihedral angles within the ring of about 8 degrees). However all optimized structures were found to be planar with fixed CS symmetry and thus were refined with fixed C2v symmetry, too.
Cyclopentadienone and fulvene (cf. Tables 5 and 6)
The initial parameters for both molecules were taken from the geometry of fulvene as determined by Baron et al. [4] and Suenram et al. [16]. For cyclopentadienone a value of 1.22 Å for the C=O bond length was used. As with the other molecules only CS symmetry was imposed and only after the calculations had converged to a planar structure these values were refined with C2v symmetry.
A survey of the obtained deviations from experimental results is given in Table 7. Reliable values for bond lengths and angles are available only for fulvene. For this molecule the MNDO method gives the best agreement with the experimental results (average deviation for all distances given in Table 6: Drrms = 0.8%, average deviation for all angles given in Table 6: Darms = 1.43%), but RHF/6-31G* (Drrms = 1.02%, Darms = 1.49%) and AM1 (Drrms = 1.08%, Da rms = 1.57%) also agree quite well. Regarding the rotational constants of fulvene, however, the situation is different: best agreement is achieved with STO-3G (average deviation of the three rotational constants: D Brms = 0.62%). AM1 (DBrms = 0.88%) and RHF/6-31G* (DBrms = 1.22%) still agree well, but MNDO constants show a larger deviation (DBrms = 1.92%).
Similar results are obtained for the other molecules where experimentally determined rotational constants are available. In three of four cases the best agreement is achieved with STO-3G, RHF/6-31G* or AM1. MINDO/3 agrees worst in these cases whereas it gives the best agreement among the semi-empirical methods for cyclopentadienone.
In summary it can be said that for the cross-conjugated compounds considered in this study, the overall agreement of the predicted rotational constants with the experimentally determined ones is quite good for STO-3G, RHF/6-31G* and AM1. Whereas the rotational constants determined by MNDO and MINDO/3 show deviations that are almost twice as high. The exact "ranking" differs depending on the molecule that is studied. Taking into account the amount of time used for the computations - several hours to days for ab initio calculations compared to some seconds to a few minutes for semi-empirical calculations - it seems that for predicting rotational constants of cross-conjugated cyclic compounds the AM1 method is the most efficient one.
The predicted constants for 2,5-cyclohexadienone, which could not be measured up to now, should be correct within ca. 1.5%. In particular the ring in the molecule should be expected to be planar.
1 P.A. Baron and R.D. Brown, Chem. Phys. 1 (1973) 444. 2 a) K. Bestle and H.-K. Bodenseh, 12th Coll. High Res. Mol. Spec. Dijon (1991) F12. b) R. Ruoff and H.-K. Bodenseh, 13th Coll. High Res. Mol. Spec. Riccione (1993) F11. 3 R.D. Brown, F.R. Burden and J.E. Kent, J. Chem. Phys. 49 (1968) 5542. 4 P.A. Baron, R.D. Brown, F.R. Burden, P.J. Domaille and J.E. Kent, J. Mol. Spectrosc., 43 (1972) 401. 5 M.-C. Lasne, J.-L. Ripoll and J.-M. Denis, Tetr.Lett. 21 (1980) 463. 6 C.S. Shiner, P.E. Vorndam and S.R. Kass, J.Am.Chem.Soc. 108 (1986) 5699. 7 H. Plieninger and W. Maier-Borst, Chem. Ber., 98 (1965) 2504. 8 J.J. Gajewski and A.M. Gortva, J. Am. Chem. Soc., 104 (1982) 334. 9 J.E. Bartmess, J. Am. Chem. Soc., 104 (1982) 335. 10 W. Hutter, H.-K. Bodenseh and A. Koch, J. Mol. Struct., 319 (1994) 73. 11 W. Hutter and H.-K. Bodenseh, J. Mol. Struct., 291 (1993) 151. 12 M. Trætteberg, P. Bakken, A. Almenningen, W. Lüttke and J. Janssen, J. Mol. Struct. 81 (1982) 87. 13 M.J. Frisch, J.S. Binkley, H.B. Schlegel, K.Raghavachari, C.F. Melius, R.L. Martin, J.J.P. Stewart, F.W. Bobrowicz, C.M. Rolfing, L.R. Kahn, D.J. Defrees, R. Seeger, R.A. Whiteside, D.J. Fox, E.M. Fleuder and J.A. Pople, GAUSSIAN86, Carnegie-Mellon Quantum Chemistry Publishing Unit, Pittsburgh PA, 1984. 14 a) M.J.S. Dewar, R.C. Bingham and D.H. Lo, J. Am. Chem. Soc. 97 (1975) 1285. b) M.J.S. Dewar and W. Thiel, J. Am. Chem. Soc. 99 (1977) 4899. c) M.J.S. Dewar, E.G. Zoebisch, E.F. Healy and J.J.P. Stewart, J. Am. Chem. Soc. 107 (1985) 3902. d) M.J.S. Dewar, J. Mol. Struct. 100 (1983) 41. 15 a) SYBYL 5.2., Tripos Inc., 1699 S. Hanley Road, Suite 303, St. Louis, MO, 63144, (1988). b) M. Clark, R.D. Cramer III and N. Van Opdenbosch, J. Comp. Chem. 10 (1989) 982. 16 R.D. Suenram and M.D. Harmony, J. Chem. Phys. (1973) 5842.
_______________________________________________________________ GAUSSIAN86 MOPAC 6-31G* STO-3G MNDO MINDO/3 AM1 _______________________________________________________________ C1=O 1.199 1.227 1.228 1.207 1.239 C1-C2 1.482 1.511 1.498 1.508 1.474 C2=C3 1.324 1.317 1.349 1.349 1.339 C3-C4 1.499 1.520 1.506 1.498 1.483 C2-H 1.075 1.083 1.092 1.106 1.102 C3-H 1.077 1.085 1.093 1.108 1.102 C4-H 1.090 1.093 1.116 1.119 1.127 C2-C1=O 121.8 122.5 122.3 122.8 122.2 C2-C1-C6 116.4 114.9 115.5 114.4 115.6 C1-C2=C3 121.7 122.3 121.9 122.5 122.0 C2=C3-C4 123.3 123.7 123.6 123.5 123.3 C3-C4-C5 113.8 113.1 113.5 113.7 113.8 C1-C2-H 116.1 116.2 117.0 118.0 115.8 C3=C2-H 122.3 121.5 121.0 119.5 122.2 C2=C3-H 120.1 120.7 121.0 121.2 121.3 C4-C3-H 116.6 115.6 115.4 115.4 115.5 C3-C4-H 109.4 109.4 109.2 110.0 109.0 H-C4-H 105.1 106.0 106.4 102.8 106.8 A 5235.36 5128.09 5174.67 5198.95 5306.56 B 2721.84 2648.84 2615.96 2620.38 2650.41 C 1810.15 1765.36 1757.00 1760.95 1788.23 µ 4.7 2.8 3.5 4.2 4.2 ________________________________________________________________*Distances are in Å, angles in degrees, rotational constants in MHz, dipole moments in Debye.
____________________________________________________________________________ GAUSSIAN86 MOPAC exp.** [11] 6-31G* STO-3G MNDO MINDO/3 AM1 ____________________________________________________________________________ C1=O 1.198 1.227 1.228 1.207 1.238 C1-C2 1.481 1.509 1.495 1.497 1.473 C2=C3 1.323 1.313 1.348 1.348 1.338 C3-C4 1.508 1.530 1.526 1.538 1.498 C2-H 1.075 1.083 1.092 1.106 1.102 C3-H 1.078 1.085 1.094 1.110 1.102 C4-C7 1.543 1.557 1.557 1.536 1.528 C7-H' 1.084 1.086 1.109 1.111 1.116 C7-H'' 1.085 1.086 1.109 1.111 1.116 C2-C1=O 122.0 122.7 122.5 123.3 122.3 C2-C1-C6 116.0 114.6 114.9 113.4 115.5 C1-C2=C3 121.6 122.4 122.3 123.0 122.2 C2=C3-C4 124.7 124.7 124.8 126.4 124.0 C3-C4-C5 111.5 111.2 110.8 107.9 112.3 C3=C2-H 122.2 121.5 120.7 118.8 122.0 C2=C3-H 119.7 120.3 119.7 119.3 121.3 C3-C4-C7 109.0 109.1 108.8 109.6 108.8 C4-C7-H' 110.8 110.6 111.2 113.5 110.4 C4-C7-H" 110.7 110.4 111.5 113.4 110.0 H'-C7-H" 108.1 108.4 107.6 105.2 108.8 H"-C7-H" 108.2 108.6 107.4 105.3 109.0 C7-C4-C8 109.3 109.1 110.8 110.4 109.2 A 3338.80 3276.33 3260.03 3306.67 3370.32 3332.16 B 1205.23 1178.95 1165.69 1151.86 1197.24 1193.07 C 1092.49 1067.37 1061.41 1050.04 1084.41 1082.21 µ 4.8 2.9 3.6 4.3 4.2 4.5 ____________________________________________________________________________*Distances are in Å, angles in degrees, rotational constants in MHz, dipole moments in Debye. **The experimental error is less than the unit of the last digit given.
____________________________________________________________________________ GAUSSIAN86 MOPAC exp.** [10] 6-31G* STO-3G MNDO MINDO/3 AM1 ____________________________________________________________________________ C1=C7 1.328 1.320 1.353 1.344 1.343 C1-C2 1.473 1.492 1.479 1.496 1.459 C2=C3 1.323 1.314 1.351 1.349 1.341 C3-C4 1.503 1.521 1.505 1.497 1.484 C2-H 1.076 1.083 1.093 1.108 1.102 C3-H 1.077 1.084 1.092 1.107 1.101 C4-H 1.090 1.093 1.116 1.120 1.127 C7-H 1.075 1.081 1.089 1.100 1.097 C2-C1=C7 122.2 122.4 122.4 123.6 122.0 C2-C1-C6 115.6 115.1 115.2 112.8 115.9 C1-C2=C3 122.4 122.6 122.7 123.8 122.2 C2=C3-C4 123.8 123.5 123.3 123.2 123.2 C3-C4-C5 112.8 112.5 112.9 113.2 113.5 C1-C2-H 117.3 116.6 117.2 116.4 116.5 C3=C2-H 120.3 120.7 120.2 119.8 121.3 C2=C3-H 119.8 120.6 121.0 120.7 121.3 C4-C3-H 116.8 115.8 115.6 116.0 115.6 C3-C4-H 109.7 109.6 109.4 110.2 109.1 H-C4-H 105.1 105.7 106.2 102.4 106.8 C1-C7-H 121.7 121.9 123.4 125.0 122.2 H-C7-H 116.7 116.2 113.1 110.0 115.6 A 5194.13 5100.58 5167.70 5181.15 5254.47 5177.82 B 2645.62 2634.14 2559.81 2523.74 2621.12 2613.15 C 1771.37 1755.54 1730.67 1714.78 1768.95 1755.84 µ 0.8 0.6 0.1 0.1 0.5 0.86 ____________________________________________________________________________*Distances are in Å, angles in degrees, rotational constants in MHz, dipole moments in Debye. **The experimental error is less than the unit of the last digit given.
____________________________________________________________________________ GAUSSIAN86 MOPAC exp. [12] 6-31G* STO-3G MNDO MINDO/3 AM1 ____________________________________________________________________________ C1=C7 1.328 1.320 1.353 1.344 1.343 1.357** C1-C2 1.472 1.491 1.477 1.493 1.459 1.478(6) C2=C3 1.323 1.314 1.350 1.348 1.339 1.352(1) C3-C4 1.512 1.530 1.525 1.539 1.498 1.493(6) C2-H 1.077 1.083 1.093 1.108 1.102 1.075(5) C3-H 1.078 1.084 1.093 1.109 1.101 1.075(5) C4-C8 1.542 1.557 1.556 1.536 1.528 1.565(3) C8-H' 1.084 1.086 1.109 1.112 1.116 1.175(4) C8-H" 1.086 1.086 1.109 1.112 1.116 1.175(4) C7-H 1.075 1.081 1.089 1.100 1.097 1.075(5) C2-C1=C7 122.4 122.6 122.7 124.1 122.1 C2-C1-C6 115.3 114.8 114.6 111.8 115.7 115.7(3) C1-C2=C3 122.4 122.7 123.1 124.3 122.4 120.8(3) C2=C3-C4 124.6 124.5 124.5 126.1 123.7 125.6** C3-C4-C5 110.7 110.8 110.3 107.5 112.0 110.5** C3=C2-H 120.3 120.7 119.8 119.1 121.1 119.0** C2=C3-H 119.5 120.3 119.8 119.0 121.4 119.5** H'-C8-H" 108.1 108.4 107.5 105.1 108.8 110.0** H"-C8-H" 108.1 108.5 107.3 105.0 108.9 110.0** C8-C4-C9 109.1 109.0 110.7 110.1 109.0 112.5§ H-C7-H 116.7 116.2 113.2 110.1 115.6 121.0**§§ d$ 0.0 0.0 0.0 0.0 0.0 7.9(6) A 3325.53 3266.14 3258.06 3300.81 3351.94 3229.59$$ B 1195.20 1185.48 1161.36 1136.31 1200.43 1183.27$$ C 1081.77 1071.21 1057.20 1035.90 1084.68 1082.05$$ µ 0.7 0.7 0.1 0.1 0.5 ____________________________________________________________________________*Distances are in Å, angles in degrees, rotational constants in MHz, dipole moments in Debye. **No error given. §Assumed. §§Calculated from complementary angle. $Dihedral angle C6-C1-C2=C3. $Calculated from the structure given by ref. [12].
____________________________________________________________________________ GAUSSIAN86 MOPAC exp.** 6-31G* STO-3G MNDO MINDO/3 AM1 ____________________________________________________________________________ C1=O 1.188 1.220 1.218 1.198 1.224 C1-C2 1.506 1.525 1.518 1.521 1.509 C2=C3 1.322 1.317 1.358 1.352 1.353 C3-C4 1.503 1.506 1.489 1.494 1.490 C2-H 1.071 1.080 1.081 1.098 1.088 C3-H 1.073 1.082 1.083 1.101 1.090 C2-C1=O 127.3 128.0 127.8 127.7 127.7 C2-C1-C5 105.5 104.0 104.4 104.6 104.7 C1-C2=C3 107.5 108.1 108.2 108.0 108.3 C2=C3-C4 109.8 109.9 109.6 109.7 109.4 C1-C2-H 123.5 123.5 123.1 124.6 122.1 C3=C2-H 129.0 128.5 128.6 127.4 129.6 C2=C3-H 127.0 127.4 127.9 128.1 128.4 C4-C3-H 123.2 122.7 122.6 122.1 122.2 A 8216.82 8170.95 8239.60 8161.31 8287.97 8152.17§ B 4024.21 3902.38 3831.37 3880.45 3843.75 3939.56§ C 2701.26 2641.04 2615.28 2629.97 2625.92 2655.80§ µ 3.6 1.9 2.8 3.4 2.9 3.1§§ ____________________________________________________________________________*Distances are in Å, angles in degrees, rotational constants in MHz, dipole moments in Debye. **The experimental error is less than the unit of the last digit given. §From ref. [2a]. §§From ref. [1].
____________________________________________________________________________ GAUSSIAN86 MOPAC exp.** 6-31G* STO-3G MNDO MINDO/3 AM1 ____________________________________________________________________________ C1=C6 1.325 1.319 1.345 1.338 1.332 1.3485(5)§ C1-C2 1.477 1.495 1.491 1.507 1.483 1.468(1)§§ C2=C3 1.333 1.324 1.366 1.359 1.363 1.357(1)§§ C3-C4 1.480 1.492 1.476 1.480 1.476 1.476(2)§ C2-H 1.073 1.080 1.082 1.100 1.088 1.078(1)§ C3-H 1.073 1.081 1.082 1.100 1.088 1.080(1)§ C6-H 1.076 1.083 1.089 1.101 1.098 1.077(1)§§ C2-C1=C6 127.3 127.8 127.8 128.5 127.4 C2-C1-C5 105.5 104.4 104.4 103.0 105.2 106.6(1)§ C1-C2=C3 108.2 108.6 109.0 109.7 108.6 107.7(1)§ C2=C3-C4 109.1 109.2 108.8 108.9 108.8 109.0(2)§ C1-C2_H 124.3 123.5 123.4 122.7 122.9 124.7(2)§ C3=C2-H 127.5 127.9 127.7 127.7 128.6 C2=C3-H 126.6 127.2 127.7 127.7 128.0 126.4(2)§ C4-C3-H 124.3 123.6 123.5 123.4 123.1 C1-C6-H 121.7 122.0 123.2 125.1 122.2 126.4(1)§§ H-C6-H 116.6 116.0 113.6 109.9 115.7 A 8240.05 8152.65 8223.85 8184.11 8218.53 8186.13§ B 3861.77 3835.81 3696.99 3654.98 3754.50 3802.73§ C 2629.45 2608.51 2550.45 2526.61 2577.17 2596.44§ µ 0.4 0.4 0.7 0.4 0.7 0.4§ ____________________________________________________________________________*Distances are in Å, angles in degrees, rotational constants in MHz, dipole moments in Debye. **If no value is given in parentheses the experimental error is less than the unit of the last digit given. §From ref. [4]]. §§From ref. [16].
_____________________________________________________________________ GAUSSIAN86 MOPAC Molecule Parameter 6-31G* STO-3G MNDO MINDO/3 AM1 _____________________________________________________________________ Distances 1.02 1.14 0.80 1.70 1.08 Fulvene Angles 1.49 1.63 1.43 1.69 1.57 Rot. const. 1.22 0.62 1.92 2.73 0.88 Cyclopentadienone Rot. const. 1.44 0.90 2.16 1.28 2.05 4,4-DMCHDO* Rot. const. 0.81 1.42 2.13 2.66 0.70 1-MCHD** Rot. const. 0.90 0.98 1.44 2.39 0.97 Average§ Rot. const. 1.12 1.02 1.93 2.34 1.27 _____________________________________________________________________*4,4-dimethyl-2,5-cyclohexadienone (IIc). **1-methylene-2,5-cyclohexadiene (IIb). §Over the rotational constants of all four molecules.
All rights reserved. No part of this service may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the publisher.No responsibility is assumed by Elsevier Science for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use of operation of any methods, products, instructions or ideas contained in the material herein.