<} }<

Quantum-Mechanical and Molecular

Mechanics Conformational Analysis
of 1,5-Cyclooctadiene

WILLIAN R. ROCHA and WAGNER B. DE ALMEIDAU

´ de Quımica
Laboratorio ´ ´
Computacional e Modelagem Molecular, Depto. Quımica, ICEx, U. F. M. G.,
Belo Horizonte, MG, 31270-901, Brazil

Received 21 November 1995; accepted 3 May 1996

ABSTRACT
The 1,5-cyclooctadiene ŽCOD. molecule can easily form complexes with transition
metals with the molecular structure of various of these complexes being proposed
with the aid of X-ray diffraction methods. The fact that the complexes exhibit
weak metal]COD bonds makes it very important in inorganic synthesis and
catalysis. In this work the potential energy surface ŽPES. for the COD molecule
was comprehensively investigated: first with molecular mechanics Žusing the
MM3 force field.; and, in a second stage, at the ab initio Hartree]Fock level of
theory employing the 3]21GU , 6]31G, and 6]31GU basis sets and also including
electron correlation effects at the Moller]Plesset second-order perturbation
theory level. This work revealed that there are three distinct conformers of the
COD molecule with the predicted lowest energy conformation being in
agreement with the proposed structure based on experimental electron
diffraction data. Q 1997 by John Wiley & Sons, Inc.

volving transition metal complexes, for example,

Introduction hydrogenation and hydroformylation of olefins.
The molecular structure of a number of such com-
plexes containing COD as the ligand were eluci-

T he 1,5-cyclooctadiene ŽCOD. compound eas-

ily complexes with transition metals. This
ligand is able to form stable complexes that exhibit
weak metal]COD bonds, which makes it very
important in inorganic synthesis and catalysis in-
This article is dedicated to the memory of our estimable
colleague Professor Milton Francisco de Jesus Filho.
* Author to whom all correspondence should be addressed.
E-mail: wagner@apolo.qui.ufmg.br
dated by the X-ray diffraction method.1 The
molecular structure of these complexes raises some
doubts about the conformation of the free ligand,
because in these complexes the COD can be bound
as a bridging ligand, assuming a chair conforma-
tion, or it can be bound as a chelate type ligand,
adopting a boat conformation.
It was proposed from an electron diffraction
study of a vapor sample of COD,2 Raman and

Journal of Computational Chemistry, Vol. 18, No. 2, 254]259 (1997)

Q 1997 by John Wiley & Sons CCC 0192-8651 / 97 / 020254-06

infrared spectra of liquid and solid COD,3 low
temperature 1 H and 13 C NMR Žnuclear magnetic
resonance. spectroscopy,4 and molecular mechan-
ics ŽMM. calculations 5 that the lowest energy con-
formation of the COD molecule is a twisted-boat
ŽTB. conformer ŽFig. 1.. It is known that the poten-
tial energy surface ŽPES. for cyclic compounds
may exhibit various minimum energy structures.
Therefore, a conformational equilibrium might be
thought to take place6 ] 8 with the corresponding
conformational interconversion path not being
generally easily determined, because the intercon-
version process may take place through a combi-
nation of various distinct internal coordinates. So
we think that an investigation of the possible sta-
ble conformers of the COD molecule can make a
contribution to our understanding of the molecular
structure of this ligand. To the best of our knowl-
edge, the only theoretical study reported so far for
the COD molecule, was within the MM calcula-
tions.5
In the present work ab initio and MM calcula-
tions were performed for several conformations of
the COD molecule. We aimed to carry out a com-
parative study to assess the reliability of the exper-
imental and theoretical approaches for the deter-
mination of the molecular structure of this cyclic
ligand. The results of this work will be used in
further studies of metal complexes involving COD
as the ligand.
Calculations
An extensive search for stationary points on the
multidimensional PES for the COD molecule was
carried out. Full unconstrained geometry opti-
mizations for several distinct trial atomic arrange-
ments led to the prediction of three distinct mini-
mum energy structures named TB, C, and SKEW
Žsee Fig. 2.. Harmonic frequency calculations were
FIGURE 1. Diagramatic representation of the
twisted-boat (TB) conformer of the 1,5-cyclooctadiene
(COD) molecule.
JOURNAL OF COMPUTATIONAL CHEMISTRY
COD QUANTUM-MECHANICAL / MM ANALYSIS
also performed to be sure that a true minimum Žall
frequencies are real. was located on the PES. The
MM calculations were performed using the MM3
force field.9 The minimum energy structures ob-
tained in the MM calculations were further fully
optimized at the ab initio Hartree]Fock ŽHF. level
of theory Žwith the harmonic frequencies also be-
ing calculated., and electron correlation was as-
sessed by the Moller]Plesset second-order pertur-
bation theory 10 ŽMP2.. All ab initio calculations
were performed with the GAMESS package.11 The
following basis set were used: 3]21GU ,12 6]31G,13
and 6]31GU .13 All calculations were carried out on
a SUN SPARC-20 workstation at the Laboratorio ´
´
de Quımica Computacional e Modelagem Molecu-
´
lar ŽLQC-MM., Departamento de Quımica, Univer-
sidade Federal de Minas Gerais ŽUFMG..
Results and Discussion
The conformations TB, C, and SKEW of the
COD molecule are depicted in Figure 2, and Table
I gives the calculated geometrical parameters. In
the last column of Table I the available experimen-
tal electron diffraction structural parameters are
quoted for comparison. From Table I it can be seen
that the experimental values reported for the w 1
w C1}C2}C3}C4x , w 2 w C2}C3}C4}C5x , w 3
w C3}C4}C5}C6x , and w6 w C8}CI}C2}C3x
angles are in good agreement with the ab initio
and MM calculated values for the global minimum
energy structure located on the PES for the COD
molecule. So, the structure observed experimen-
tally is unambiguously assigned to the TB con-
former. It is known that the MM approach yields
good molecular geometry results for this kind of
molecular system Žhydrocarbons.14 and the good
agreement with the ab initio and experimental re-
sults come as no surprise. The other calculated
structural parameters, bond distances and bond
angles, for the TB conformer are consistent with
the values estimated experimentally. These param-
eters are not very sensitive to conformational
changes. Also the quality of the basis set used in
the ab initio calculations does not appear to signifi-
cantly affect the geometry results.
In the experimental electron diffraction study
the authors could not rule out the possible pres-
ence of a second conformer for the COD molecule,
named C, up to 10%. The theoretical study re-
ported here provides support for the possibility of
the existence of more than one conformer, because
255
ROCHA AND DE ALMEIDA

FIGURE 2. The three minimum energy structures on the PES for the COD molecule: (a) TB conformer (global
minimum), (b) C conformer, and (c) SKEW conformer.

three true minimum energy structures were lo-
cated on the PES for the COD molecule. The MP2
zero point energy ŽZPE. corrected energy differ-
ence of 2 kcalrmol Žsee Table II. between struc-
tures TB and C suggests that if the energy barrier
for interconversion is not very high, both isomers
may coexist in the gas phase mixture. However,
we were not able to find a first-order transition
state ŽTS. structure connecting the TB and C con-
formers on the multidimensional PES. Also the

256 VOL. 18, NO. 2

 COD QUANTUM-MECHANICAL / MM ANALYSIS
TABLE I.
Structural Parameters for COD Molecule.

MM3 HF / 3]21GU HF / 6]31G HF / 6]31GU Expt.a

TB Conformer
R (C1}H9) 1.104 1.076 1.077 1.079 1.073 (4)
R (C1}C2) 1.338 1.318 1.327 1.322 1.336 (3)
R (C2}C3) 1.509 1.514 1.508 1.510 1.513 (5)
R (C3}C4) 1.534 1.549 1.541 1.536 1.536 (9)
R (C5}C6) 1.338 1.318 1.327 1.322 }
R (C6}C7) 1.509 1.514 1.508 1.510 }
R (C7}C8) 1.530 1.549 1.541 1.536 }
u1 (C2}C1}H9) 114.9 117.2 116.5 116.3 }
u 2 (C4}C3}H14) 111.3 110.7 110.3 110.4 }
u 3 (C1}C2}C3) 129.6 130.1 130.6 130.8 129.3
u4 (C2}C3}C4) 118.0 117.1 117.9 118.1 112.4
u 5 (C3}C4}C5) 112.7 112.3 113.8 113.9 116.8
u 6 (C4}C5}C6) 127.6 128.4 129.1 129.1 131.4
u 7 (C5}C6}C7) 129.6 130.1 130.6 130.8 }
u 8 (C6}C7}C8) 118.0 117.1 117.9 118.1 }
u 9 (C7}C8}C1) 112.7 112.3 113.8 113.9 }
u10 (C8}C1}C2) 127.6 128.4 129.1 129.1 }
v1 (C1}C2}C3}C4) 20.0 20.1 15.6 15.2 8.1 (44)
v 2 (C2}C3}C4}C5) 56.7 57.6 60.3 60.3 63.8 (71)
v 3 (C3}C4}C5}C6) y87.1 y85.7 y82.9 y82.6 y86.5 (81)
v4 (C4}C5}C6}C7) 2.15 0.2 0.0 0.2 }
v 5 (C5}C6}C7}C8) 20.0 20.1 15.6 15.2 }
v 6 (C8}C1}C2}C3) 2.15 0.2 0.0 0.2 6.2 (66)
v 7 (C1}C8}C7}C6) 56.7 57.6 60.3 60.3 }
v 8 (C2}C1}C8}C7) y87.1 y85.7 y82.3 y82.6 }
C Conformer
R (C1}H9) 1.105 1.076 1.077 1.078
R (C1}C2) 1.336 1.318 1.326 1.321
R (C2}C3) 1.515 1.516 1.512 1.514
R (C3}C4) 1.543 1.568 1.557 1.552
R (C5}C6) 1.336 1.318 1.326 1.321
R (C6}C7) 1.515 1.516 1.512 1.514
R (C7}C8) 1.515 1.568 1.557 1.552
u1 (C2}C1}H9) 117.7 118.3 117.5 117.2
u 2 (C4}C3}H14) 109.2 109.0 108.7 108.8
u 3 (C1}C2}C3) 127.0 126.1 127.1 127.5
u4 (C2}C3}C4) 113.0 112.5 114.1 114.1
u 5 (C3}C4}C5) 113.0 112.5 114.1 114.1
u 6 (C4}C5}C6) 127.0 126.1 127.1 127.5
u 7 (C5}C6}C7) 127.0 126.1 127.1 127.5
u 8 (C6}C7}C8) 113.0 112.5 114.1 114.1
u 9 (C7}C8}C1) 113.0 112.5 114.1 114.1
u10 (C8}C1}C2) 127.0 126.1 127.1 127.5
v1 (C1}C2}C3}C4) y71.3 y72.4 y70.3 y70.0
v 2 (C2}C3}C4}C5) y98.2 y100.7 y97.3 y96.5
v 3 (C3}C4}C5}C6) 71.3 72.4 70.3 70.0
v4 (C4}C5}C6}C7) 0.0 0.0 0.0 0.0
v 5 (C5}C6}C7}C8) y71.3 y72.4 y70.3 y70.0
v 6 (C8}C1}C2}C3) 0.0 0.0 0.0 0.0
v 7 (C1}C8}C7}C6) 98.2 100.8 97.4 96.5
v 8 (C2}C1}C8}C7) y71.3 y72.4 y70.3 y70.0
SKEW conformer
R (C1}H9) 1.105 1.076 1.078 1.077
R (C1}C2) 1.335 1.318 1.327 1.327

JOURNAL OF COMPUTATIONAL CHEMISTRY 257

ROCHA AND DE ALMEIDA

TABLE I.
(continued)

MM3 HF / 3]21GU HF / 6]31G HF / 6]31GU Expt.a

R (C2}C3) 1.512 1.517 1.511 1.512 y86.5 (81)

R (C3}C4) 1.540 1.557 1.548 1.548
R (C5}C6) 1.335 1.318 1.326 1.327
R (C6}C7) 1.512 1.517 1.512 1.511
R (C7}C8) 1.540 1.557 1.548 1.548
u1 (C2}C1}H9) 116.7 117.3 116.7 116.6
u 2 (C4}C3}H14) 110.7 109.2 109.2 109.2
u 3 (C1}C2}C3) 129.0 128.1 129.0 129.0
u4 (C2}C3}C4) 115.8 115.2 116.5 116.5
u 5 (C3}C4}C5) 115.0 115.2 116.5 116.9
u 6 (C4}C5}C6) 129.0 128.1 129.0 129.6
u 7 (C5}C6}C7) 129.0 128.1 129.0 129.6
u 8 (C6}C7}C8) 115.8 115.2 116.5 116.9
u 9 (C7}C8}C1) 115.8 115.2 116.5 116.9
u10 (C8}C1}C2) 129.0 128.1 129.0 129.6
v1 (C1}C2}C3}C4) 43.4 44.4 43.2 43.1
v 2 (C2}C3}C4}C5) y96.6 y99.9 y96.0 y94.1
v 3 (C3}C4}C5}C6) 43.4 44.6 43.3 42.6
v4 (C4}C5}C6}C7) 3.2 2.4 2.7 3.1
v 5 (C5}C6}C7}C8) 43.4 44.4 43.1 43.1
v 6 (C8}C1}C2}C3) 3.23 2.4 2.7 2.8
v 7 (C1}C8}C7}C6) y96.6 y99.9 y96.0 y94.1
v 8 (C2}C1}C8 ]C7) 43.4 44.6 43.3 42.6
Angles in degrees and bond distances in angstroms.
a
Electron dibraction data.2 The experimental uncertainties given in parentheses are 2 s and include estimates of correlation and
systematic error.

population ratio of less than about 10% of C in
relation to TB, which was not ruled out in the
electron diffraction study, appears to be reason-
able. Our respective population ration value at the
temperature of the electron diffraction experiment,2
evaluated with the MP2r6]31GUrr6]31GU ZPE
corrected energy difference value using the Boltz-
mann distribution law, is 5%; and the relative
population between the TB and SKEW conformers
is negligible Ž0.3%..
The relative energies of the three possible con-
formers for the COD molecule are given in Table
II. The double slash means that a single point
calculation at the fully optimized HF geometry

TABLE II.
Energy Differences (kcal / mol) in Relation to Lowest Energy Conformer for COD Molecule.

TB C SKEW

MM3 0.0 4.530 6.010

HF / 3 ]21GU 0.0 2.015 3.495
HF / 6 ]31G 0.0 1.174 (0.840) a 3.743 (3.538) a
HF / 6 ]31GU 0.0 2.868 (2.662) a 4.064 (3.730) a
MP2 / 6 ]31G // 6 ]31G b 0.0 1.578 (1.244) a 3.751 (3.417) a
MP2 / 6 ]31GU // 6 ]31G b 0.0 2.237 (1.903) a 4.132 (3.927) a
MP2 / 6 ]31GU // 6 ]31GUb 0.0 2.211 (2.006) a 4.215 (3.881) a
a
Calculated using HF / 6 ]31G zero point corrected energies.
b
Single point calculation.

258 VOL. 18, NO. 2


was performed. It can be seen that the energy
differences predicted by the MM and ab initio
calculations led to the prediction of the TB con-
former as the global minimum on the PES and also
structure C as the lowest energy local minimum.
The chart below shows the energy order predicted
by the MM and ab initio calculations and by the
experiment. The AM115 and PM316 semiempirical
energy results are also shown for comparison. At
the AM1 and PM3 semiempirical level the C con-
former is more stable than the TB structure by
1.875 and 0.817 kcalrmol, respectively. So it can be
seen that the AM1 and PM3 energy differences are
lower than the ab initio and MM3 values with the
energy order being in complete disagreement with
experimental and other theoretical values.
Stability Order Method
TB ) C ) SKEW MM3
C ) SKEW ) TB AM1
C ) TB ) SKEW PM3
TB ) C ) SKEW HF / 3]21G*
TB ) C ) SKEW HF / 6]31G
TB ) C ) SKEW HF / 6]31G*
TB ) C ) SKEW MP2 / 6]31G // 6]31G
TB ) C ) SKEW MP2 / 6]31G* // 6]31G
TB ) C ) SKEW MP2 / 6]31G* // 6]31G*
TB ) C Expt. 2
As seen from the chart above, the energy trend
is not affected as the basis set is increased and
electron correlation effects are included. The re-
sults reported in Table I show that the addition of
a polarization function to the 6]31G basis set does
not produce a significant change in the geometrical
parameters. However, the relative energies are
considerably affected Žsee Table II.. The MP2r 6]
31GUrr6]31GU relative energy is lower than the
HFr 6]31GU one for the C conformer and higher
for the SKEW conformer. Therefore, it does not
appear to always be generally true that the energy
differences evaluated at the HF level are overesti-
mated.
Finally, an important issue that emerges from
the present study is that the HF level of theory
with a moderate basis set is sufficient for the
prediction of the energetic order of the three con-
formers of the COD molecule. However, the addi-
tion of polarization functions to the basis set and
inclusion of electron correlation effects are very
JOURNAL OF COMPUTATIONAL CHEMISTRY
COD QUANTUM-MECHANICAL / MM ANALYSIS
important for an accurate prediction of relative
energies.
Acknowledgments
W.R.R. would like to thank the CNPq ŽConselho
´
Nacional de Desenvolvimento Cientıfico e Tecno-
´
logico. for a research grant. W.B.D.A. also thanks
the CNPq for support.
References
1. Ža. J. A. Ibers and R. G. Snyder, Acta Crystallogr., 15, 923
Ž1962.; Žb. J. H. van der Hender and W. C. Baird, J. Am.
Chem. Soc., 85, 1009 Ž1963.; Žc. M. D. Glick and L. F. Dahl, J.
Organomet. Chem., 3, 200 Ž1965..
2. K. Hagner, L. Hedberg, and K. Hedberg, J. Phys. Chem., 86,
117 Ž1982..
3. P. J. Hendra and D. B. Powell, Spectrochim. Acta, 17, 913
Ž1961..
4. F. A. L. Anet and L. Kozerski, J. Am. Chem. Soc., 95, 3407
Ž1973..
5. Ža. N. L. Allinger and J. T. Sprague, Tetrahedron, 31, 21
Ž1975.; Žb. O. Ermer, J. Am. Chem. Soc., 98, 3964 Ž1976..
6. A. M. G. Do Val, A. C. Guimaraes, ˜ and W. B. De Almeida,
J. Heterocyc. Chem., 32, 557 Ž1995..
7. W. B. De Almeida, A. M. G. Do Val, H. F. Dos Santos, and
A. C. Guimaraes, ˜ In Annals of The Third Latin American
Conference on Physical Organic Chemistry, Florianopolis, ´
Brazil, Ž1996. in press.
8. H. F. Dos Santos, A. M. G. Do Val, A. C. Guimaraes, ˜ and W.
B. De Almeida, Inf. Tecnol. to appear.
9. N. L. Allinger and Y. Fan, J. Comput. Chem., 14, 655 Ž1993.,
and references therein.
10. C. Moller and M. S. Plesset, Phys. Rev., 46, 618 Ž1934..
11. M. W. Schimidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert,
M. S. Gordon, J. H. Jensen, S. Kosek, N. Matsunaga, K. A.
Nguyen, S. J. Su, T. L. Windus, M. Dupuis, and J. A.
Montgomery, J. Comput. Chem., 14, 1347 Ž1993..
12. S. Huzinaga, J. Andzelm, M. Klobukowski, E. Radzio]And-
zelm, Y. Sakai, and H. Tatewaki, Gaussian Basis Sets for
Molecular Calculations, Elsevier, Amsterdam, 1984.
13. Ža. R. Ditchfield, W. J. Hehre, and J. A. Pople, J. Chem.
Phys., 54, 724 Ž1971.; Žb. W. J. Hehre, R. Ditchfield, and J. A.
Pople, J. Chem. Phys., 56, 2257 Ž1972..
14. Ža. N. L. Allinger, Y. H. Yuh, and J. H. Lii, J. Am. Chem.
Soc., 111, 111 Ž1989.; Žb. J. H. Lii and N. L. Allinger, J. Am.
Chem. Soc., 111, 8566 Ž1989.; Žc. N. L. Allinger, F. Li, and L.
Yan, J. Comput. Chem., 11, 848 Ž1990..
15. M. J. S. Dewar, E. G. Zoebisch, E. F. Healy, and J. J. P.
Stewart, J. Am. Chem. Soc., 107, 3902 Ž1985..
16. Ža. J. J. P. Stewart, J. Comput. Chem., 10, 209 Ž1989.; Žb. J. J. P.
Stewart, J. Comput. Chem., 10, 221 Ž1989..
259