Molecular Structure of 3,4-Dimethylenehexa-1,5-diene ([4]Dendralene), C8H10, in the Gas Phase As Determined by Electron Diffraction and ab Initio Calculations

Article abstract of DOI:10.1021/jo962091h

The molecular structure of 3,4-dimethylenehexa-1,5-diene ([4]dendralene), C₈H₁₀, has been determined in the gas phase. A single conformer with C₂ symmetry, having two almost planar, anti butadiene groups orientated with a dihedral angle C(2)C(3)C(4)C(5) of 71.7(19) °, is detected by electron diffraction employing flexible restraints derived from ab initio computations. Other experimental structural parameters (r_α/pm, ∠_α/°) are: C(1)=C(2) 133.4(1), C(3)=C(7) (not in main chain) 134.0(1), C(2)-C(3) 147.4(2), C(3)-C(4) 149.6(3), C(1)C(2)C(3) 124.4(3), C(2)C(3)C(4) 119.2(5), C(4)C(3)C(7) 117.6(7), and C(7)C(3)C(2)C(1) -174.8(28). Ab initio computations at the MP2/6-311G* level predict that the vapor consists of ca. 90percent of the conformer found experimentally, the other 10percent comprising four other conformers.

Full text of DOI:10.1021/jo962091h

J . Org. Chem. 1997, 62, 2767-2773
2767
Molecu la r Str u ctu r e of 3,4-Dim eth ylen eh exa -1,5-d ien e
([4]Den d r a len e), C8H10, in th e Ga s P h a se As Deter m in ed by
Electr on Diffr a ction a n d a b In itio Ca lcu la tion s
Paul T. Brain, Bruce A. Smart, Heather E. Robertson, Martin J . Davis,†
David W. H. Rankin,* William J . Henry, and Ian Gosney
Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3J J , U.K.
Received November 8, 1996X
The molecular structure of 3,4-dimethylenehexa-1,5-diene ([4]dendralene), C8H10, has been
determined in the gas phase. A single conformer with C2 symmetry, having two almost planar,
anti butadiene groups orientated with a dihedral angle C(2)C(3)C(4)C(5) of 71.7(19)°, is detected
by electron diffraction employing flexible restraints derived from ab initio computations. Other
experimental structural parameters (rR/pm, R/°) are: C(1)dC(2) 133.4(1), C(3)dC(7) (not in main
chain) 134.0(1), C(2)-C(3) 147.4(2), C(3)-C(4) 149.6(3), C(1)C(2)C(3) 124.4(3), C(2)C(3)C(4) 119.2(5),
C(4)C(3)C(7) 117.6(7), and C(7)C(3)C(2)C(1) -174.8(28). Ab initio computations at the MP2/6-
311G* level predict that the vapor consists of ca. 90% of the conformer found experimentally, the
other 10% comprising four other conformers.
Sch em e 1
In tr od u ction
Cross-conjugated acyclic polyenes, otherwise known as
dendralenes, are interesting compounds from a synthetic
and a theoretical point of view.1 As the simplest member
of this class of compounds, [3]dendralene 1 (3-methyl-
enepenta-1,4-diene) has been the subject of numerous
theoretical studies,2 but relatively little experimental
work.3 The molecular structure of 1 has been investi-
gated by gas-phase electron diffraction (GED) to reveal
an anti,skew conformation (Scheme 1) consisting of an
anti-butadiene fragment with the other vinyl group
subtending a dihedral angle of ca. 40° with the butadiene
plane, rotated from the syn configuration.4
Sch em e 2
In the present paper, we report the results of a study
of the molecular structure of [4]dendralene 2 (3,4-
dimethylenehexa-1,5-diene) by GED and ab initio com-
putations. A priori, [4]dendralene is a dehydrodimer of
butadiene, having two overlapping cross-conjugated diene
systems. Although 2 was first prepared in 1962,5 details
of its structure remain unknown, not least since it proved
difficult to obtain in a pure state until 1985.6 The
similarity of the UV spectrum (λmax ) 216.5 nm)5,7 to that
of butadiene (λmax ) 217 nm) suggests that [4]dendralene
may be regarded as two butadienyl groups bonded
together without significant further cross conjugation.
Different arrangements of the four double bonds give
rise to six possible conformers, the planar forms of which
are shown in Scheme 2. Each conformer, planar or
nonplanar, is described by reference to the conformation
of each butadiene, and to the orientation of the vinyl
groups relative to one another. In this study, we have
investigated the relative populations of these conformers
and their distortions from the idealized structures of
Scheme 2.
† Present address: DFEI/SME, University of Cranfield, Cranfield,
Beds., MK43 0AL, UK.
X Abstract published in Advance ACS Abstracts, April 1, 1997.
(1) For a review of cross-conjugated molecules of the dendralene-
type, see: Hopt, H. Angew. Chem., Int. Ed. Engl. 1984, 23, 948.
(2) See: Norinder, V. J . J . Mol. Struct. (THEOCHEM) 1987, 150,
85 and references therein.
(3) (a) Cadogan, J . I. G.; Cradock, S.; Gillam, S; Gosney, I. J . Chem.
Soc., Chem. Commun. 1991, 114. (b) Trahanovsky, W. S.; Koeplinger,
K. A. J . Org. Chem. 1992, 57, 4711.
(4) Almenningen, A.; Gatial, A.; Grace, D. S. B.; Hopf, H.; Klaeboe,
P.; Lehrich, F.; Nielsen, C. J .; Powell, D. L.; Trætteberg, M. Acta Chem.
Scand. 1988, A42, 634.
Exp er im en ta l Section
(5) [4]Dendralene 2 was first prepared as part of a complex mixture
of at least 20 products from the pyrolysis of the tetraacetate of butane-
1,2,3,4-tetracarboxylic acid. See: Bailey, W. J .; Nielsen, N. A. J . Org.
Chem. 1962, 27, 3088. Other methods for its preparation are reviewed
in ref 1, but a similar lack of specificity and the synthetic effort required
to make the different precursors renders them essentially useless for
practical purposes.
Syn t h esis. [4]Dendralene was prepared by extrusion of
SO2 from 6,7-dimethylene-3-thiabicyclo[3.2.0]heptane-3,3-
dioxide (3) under flash vacuum pyrolytic conditions at 823 K
(Scheme 3).6 The general technique and apparatus for flash
vacuum pyrolysis have been described previously.8
(6) Buchan, C. M.; Cadogan, J . I. G.; Gosney, I.; Henry, W. J . J .
Chem. Soc., Chem. Commun. 1985, 1785.
(7) Bee, L. K.; Everett, J . W.; Garratt, P. J . Tetrahedron 1977, 33,
2143.
(8) Cadogan, J . I. G.; Cameron, D. K.; Gosney, I.; Tinley, E. J .; Wyse,
S. J .; Amaro, A. J . Chem. Soc., Perkin Trans. 1 1991, 2081.
S0022-3263(96)02091-9 CCC: $14.00 © 1997 American Chemical Society

2768 J . Org. Chem., Vol. 62, No. 9, 1997
Brain et al.
Ta ble 1. Nozzle-to-P la te Dista n ces, Weigh tin g F u n ction s, Cor r ela tion P a r a m eter s, Sca le F a ctor s a n d Electr on
Wa velen gth s
weighting function/nm-1
nozzle-to-plate distance/mm
∆s
smin
sw1
sw2
smax
correlation parameter
scale factor ka
electron wavelength/pmb
285.4
128.3
2
4
20
40
40
60
124
280
144
336
-0.1374
-0.0901
0.772(3)
0.781(7)
5.686
5.687
a
b
Figures in parentheses are the estimated standard deviations. Determined by reference to the scattering pattern of benzene vapor.
1000 workstation using the Gaussian suite of programs.14
Sch em e 3
Geometry optimizations were performed using standard gradi-
ent techniques at the SCF level of theory using the 3-21G*,15-17
6-31G*,18-20 and 6-311G*21 basis sets. Subsequently, the two
larger basis sets were adopted for calculations at the MP2-
(FC) level of theory.
Vibrational frequency calculations were performed on all
stationary points located at the SCF/3-21G* and SCF/6-31G*
levels to verify structures as minima on the potential-energy
surface (PES). In addition, a geometry optimization and
frequency calculation were undertaken for the lowest-energy
Typically, 0.155 g (0.900 mmol) of 3,6 admixed with galvi-
noxyl to prevent polymerization, were pyrolyzed (inlet, 358 K;
oven, 823 K) at 9 × 10-3 mmHg pressure. After ca. 90 min,
conformer using density-functional methods at the B3LYP/6-
all of the starting sulfone had evaporated and a colorless band
31G* level.22 A fine grid containing 99 radial shells and 302
had collected in the trap held at 77 K. After completion of
angular points per shell for each atom was used for these
the pyrolysis, the system was isolated from the pump and filled
calculations.
with nitrogen gas. Isolation of pure [4]dendralene 2 as a clear
liquid (0.080 g, 83%) was achieved by removal of SO2 from the
crude product mixture by pumping on the trap warmed to 233
K at 0.05 mmHg pressure, followed by flash distillation,
allowing the trap to warm to room temperature. The purity
of the compound was checked by reference to the 1H and 13C
NMR spectra of a CDCl3 solution: 1H NMR δ 6.43 (2H, dd, J
) 17.30 and 10.50 Hz), 5.24 (2H, m), 5.18 (2H, d, J ) 17.3
Hz), 5.10 (2H, d, J ) 10.5 Hz), 5.05 (2H, m) and identical to
data reported by Roth et al.9 to whom we are indebted for
providing a copy of the spectrum; 13C(DEPT) NMR δ 137.30,
117.47 and 116.33.
Electr on -Diffr a ction Mea su r em en ts. Electron-scatter-
ing intensities were recorded on Kodak Electron Image plates
using the Edinburgh gas-diffraction apparatus operating at
ca. 44.5 kV (electron wavelength ca. 5.7 pm).10 Nozzle-to-plate
distances were ca. 128 and 285 mm, yielding data in the s
range 20-336 nm-1; four usable plates were obtained at each
distance. The sample and nozzle were held at ca. 293 K during
the exposure periods.
The scattering patterns of benzene were also recorded for
the purpose of calibration; these were analyzed in exactly the
same way as those of the [4]dendralene so as to minimize
systematic errors in the wavelengths and camera distances.
Nozzle-to-plate distances, weighting functions used to set up
the off-diagonal weight matrix, correlation parameters, final
scale factors, and electron wavelengths for the measurements
are collected together in Table 1.
The electron-scattering patterns were converted into digital
form using a computer-controlled J oyce-Loebl MDM6 mi-
crodensitometer with a scanning program described previ-
ously.11 The programs used for data reduction11 and least-
squares refinement12 have been described elsewhere; the
complex scattering factors employed were those listed by Ross
et al.13
Molecu la r Mod el
Several different models were used to generate the
atomic coordinates of [4]dendralene employing the atom-
numbering scheme shown in Figure 1. For a single
conformer with C2 symmetry, twelve geometrical param-
eters were used to define the molecular structure, as
given in Table 2. These include five bond distances: the
average of all CsC and CdC distances (p1); the difference
between the mean CsC and the mean CdC distances
(p2); the difference between the CsC distances (p3); the
difference between the CdC distances (p4); and the mean
of the three different CsH distances (p5). For each
distinct type of CsH distance, the difference from the
mean was fixed at the theoretical value (MP2/6-31G*
level). Assuming planarity of the atoms within the
groups H(2)C(2)C(1)H2 and C(2)C(3)C(7)H2, the other
parameters consisted of the CdCsH(1) angle (p6), as-
sumed equal for all CdCsH2 groups; the vinylic CsCdC
angle (p7); the angle H(2)sC(2)dC(1); the torsion of the
vinyl group about C(2)sC(3), i.e. the dihedral angle
C(7)C(3)C(2)C(1), defined as 0° for a syn planar confor-
mation (p9); the angles C(4)sC(3)sC(2) (p10) and C(4)sC-
(3)dC(7) (p11); and the orientation of one butadiene group
(14) Gaussian 94 (Revision C.2). Frisch, M. J .; Trucks, G. W.;
Schlegel, H. B.; Gill, P. M. W.; J ohnson, B. G.; Robb, M. A.; Cheesman,
J . R.; Keith, T. A.; Petersson, G. A.; Montgomery, J . A.; Raghavachari,
K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.;
Cioslowski, J .; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.;
Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J . L.;
Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.;
Defrees, D. J .; Baker, J .; Stewart, J . P.; Head-Gordon, M.; Gonzalez
C.; Pople, J . A. Gaussian Inc., Pittsburgh, PA, 1995.
Th eor etica l Ca lcu la tion s. A graded series of ab initio
molecular-orbital calculations was undertaken to predict geo-
metrical parameters and to obtain theoretical harmonic force
fields. All calculations were carried out on a DEC Alpha APX
(15) Binkley, J . S.; Pople J . A.; Hehre, W. J . J . Am. Chem Soc. 1980,
102, 939.
(16) Gordon, M. S.; Binkley, J . S.; Pople, J . A.; Pietro, W. J .; Hehre,
W. J . J . Am. Chem. Soc. 1982, 104, 2797.
(17) Pietro, W. J .; Francl, M. M.; Hehre, W. J .; Defrees, D. J .; Pople,
J . A.; Binkley, J . S. J . Am. Chem. Soc. 1982, 104, 5039.
(18) Hehre, W. J .; Ditchfield R.; Pople, J . A. J . Chem. Phys. 1972,
56, 2257.
(19) Hariharan P. C.; Pople, J . A. Theor. Chim. Acta 1973, 28, 213.
(20) Gordon, M. S. Chem. Phys. Lett. 1980, 76, 163.
(21) Krishnan, R.; Binkley, J . S.; Seeger R.; Pople, J . A. J . Chem.
Phys. 1980, 72, 650.
(9) Roth, W. R.; Scholz, B. P.; Breuckmann, R.; J elich, K.; Lennartz,
H.-W. Chem. Ber. 1982, 115, 1934.
(10) Huntley, C. M.; Laurenson, G. S.; Rankin, D. W. H. J . Chem.
Soc., Dalton Trans. 1980, 954.
(11) Cradock, S.; Koprowski, J .; Rankin, D. W. H. J . Mol. Struct.
1981, 77, 113.
(12) Boyd, A. S. F.; Laurenson, G. S.; Rankin, D. W. H. J . Mol.
Struct. 1981, 71, 217.
(13) Ross, A. W.; Fink, M.; Hilderbrandt, R. In International Tables
for Crystallography; Wilson, A. J . C., Ed.; Kluwer Academic Publish-
ers: Dordrecht, The Netherlands, Boston, MA, and London, 1992; Vol.
C, p 245.
(22) For example, see: Downs, A. J .; Greene, T. M.; Harman, L. A.;
Souter, P. F.; Brain, P. T.; Pulham, C. R.; Rankin, D. W. H.; Robertson,
H. E.; Hofmann, M.; Schleyer, P. v. R. Inorg. Chem. 1995, 34, 1799.

Gas-Phase Structure of [4]Dendralene
J . Org. Chem., Vol. 62, No. 9, 1997 2769
a
b
c
d
e
F igu r e 1. Views of [4]dendralene: (a) conformer (1) in the optimum refinement of the electron-diffraction data: (i) perspective
view, and (ii) view along C(3)-C(4); (b-e) perspective views of conformers (2-5), respectively, optimized at the MP2/6-311G*
level of theory.
Ta ble 2. Molecu la r P a r a m eter s (r r° str u ctu r e)
no.
parametera
distance/pm or angle/°b
1
2
3
4
5
6
7
8
9
1/7[rC(3)-C(4) + 2rC(2)-C(3) + 2rC(1)-C(2) + 2rC(3)-C(7)]
139.87(4)
14.4(1)
2.2(5)
1/3[rC(3)-C(4) + 2rC(2)-C(3)] - 1/2[rC(1)-C(2) + rC(3)-C(7)]
rC(3)-C(4) - rC(2)-C(3)
rC(3)-C(7) - rC(1)-C(2)
1/5[2rC(1)-H(1) + rC(2)-H(2) + 2rC(7)-H(7)]
C(2)C(1)H(1)
C(3)C(2)C(1)
H(2)C(2)C(1)
C(7)C(3)C(2)C(1)
C(4)C(3)C(2)
C(4)C(3)C(7)
C(2)C(3)C(4)C(5)
0.5(3)
109.1(1)
119.4(4)
124.4(3)
121.4(7)
-174.8(28)
119.2(5)
117.6(7)
71.7(19)
10
11
12
a
b
For atom-numbering scheme, see Figure 1. Figures in parentheses are the estimated standard deviations.
relative to the other defined by the torsion angle C(2)sC-
(3)sC(4)sC(5) (p12).
6-31G*, and MP2/6-31G* levels failed to locate a station-
ary point derived from starting position (6), i.e. anti,anti
(anti); this arrangement was found always to collapse to
a structure derived from starting position (1), anti,anti
(syn).
On the basis of the ab initio calculations detailed below,
several other conformers are also expected to be present
in small amounts (<10%) in the gas phase at room
temperature. Thus, other models were employed which
explored the possibility of contributions to the scattering
by these conformers. Since the fraction of each of these
conformers was predicted to be small, their geometrical
parameters were not expected to be refinable. Thus,
their structures were defined in terms of bond-distance
and bond-angle differences, fixed relative to the predomi-
nant C2 conformer, and fixed torsion angles; these angles
and differences were fixed at the values computed ab
initio (MP2/6-31G* level).
The relative energies of the five local minima are
presented in Table 3. At all levels, conformer (1) was
found to be the lowest energy structure by at least 6.8
kJ mol-1. The energy ordering of the conformers did not
alter with improvements in the theoretical treatment
except in the case of conformers (2) and (3), for which
the energy separation was never more than 0.7 kJ mol-1
when the effects of zero-point energy are taken into
consideration. At the highest level of theory available
to us [MP2/6-311G* + 0.9‚ZPE(SCF/6-31G*)], the energy
ordering of the conformers is predicted to be (1) < (3) <
(2) < (4) < (5). Neglecting entropy effects, and taking
account of the two-fold degeneracy of conformers (3) and
(4), this corresponds to an equilibrium mixture consisting
of 89.1% (1), 5.7% (3), 2.6% (4) and 2.6% (2) [<0.1% (5)]
in the gas phase at 293 K.
Resu lts
Ab In itio Ca lcu la tion s. Each of the six possible
conformers of [4]dendralene was considered for theoreti-
cal studies. Vibrational-frequency calculations performed
at the SCF/3-21G* and SCF/6-31G* levels showed that
conformers derived from the starting positions 1-5 in
Scheme 2 all represent local minima on the PES. An
extensive search of the surface at the SCF/3-21G*, SCF/
Selected geometrical parameters for the lowest energy
conformer (1), optimized at several different theoretical
levels, are reported in Table 4. Bond lengths were found
to be fairly insensitive to the adopted theoretical method,

2770 J . Org. Chem., Vol. 62, No. 9, 1997
Brain et al.
Ta ble 3. Va r ia tion of Rela tive En er gy (k J m ol-1) a n d P r op or tion w ith a b In itio Level for Con for m er s of [4]Den d r a len e
relative energya (% at 293 K)
level/basis set
SCF/3-21G*
SCF/6-31G*
MP2/6-31G*
SCF/6-311G*
MP2/6-311G*
conformer (1)
conformer (2)
conformer (3)
conformer (4)
conformer (5)
0.0 (83.07)
0.0 (87.50)
0.0 (82.37)
0.0 (90.82)
0.0 (89.05)
7.46 (3.89)
7.90 (3.42)
6.86 (4.93)
8.90 (2.35)
8.63 (2.58)
6.79 (10.23)
7.96 (6.67)
7.48 (7.64)
8.55 (5.43)
8.37 (5.73)
9.97 (2.77)
10.46 (2.39)
8.50 (5.03)
11.90 (1.37)
10.27 (2.63)
18.94 (0.03)
20.20 (0.02)
19.33 (0.03)
21.01 (0.02)
21.47 (0.01)
∆ZPE (SCF/3-21G*)b
∆ZPE (SCF/6-31G*)b
0.00
0.00
0.42
0.71
-0.08
0.00
0.17
0.62
-1.07
-0.78
a
Corrected for zero-point energy scaled by 0.9; SCF/3-21G* corrected for ZPE(SCF/3-21G*) and all others for ZPE(SCF/6-31G*).
Unscaled. Relative to the ZPE for conformer (1).
b
Ta ble 4. Va r ia tion of Str u ctu r a l P a r a m eter s (r /p m , /°) w ith a b In itio Level for Con for m er (1)
theoretical level/basis set (re)
parametera
C(2)sC(3)
C(3)sC(4)
C(1)dC(2)
SCF/3-21G*
SCF/6-31G*
SCF/6-311G* B3LYP/6-31G*
MP2/6-31G*
MP2/6-311G* expt (rR°) GEDb
147.8
150.0
131.9
132.3
107.3
121.6
125.3
119.7
147.9
150.0
132.2
132.6
107.6
121.6
126.1
119.0
-173.3
119.3
120.5
77.8
147.9
150.0
132.1
132.5
107.6
121.6
126.2
118.9
-174.1
119.3
120.5
79.1
147.1
150.1
133.9
134.5
108.7
121.6
126.1
119.0
-174.4
119.5
120.2
76.3
146.8
149.1
134.4
134.9
108.7
121.4
124.8
119.3
-172.4
119.1
120.1
73.8
147.0
149.2
134.5
135.1
108.6
121.3
124.6
119.3
-173.4
118.9
120.1
74.0
147.4(2)
149.6(3)
133.4(1)
134.0(1)
109.1(1)
119.4(4)
124.4(3)
121.4(7)
-174.8(28)
119.2(5)
117.6(7)
71.7(19)
C(3)dC(7)
C-H (mean)
CdCH(H) (mean)
C(3)C(2)C(1)
H(2)C(2)C(1)
C(7)C(3)C(2)C(1) -174.3
C(4)C(3)C(2)
118.4
120.5
80.2
C(4)C(3)C(7)
C(2)C(3)C(4)C(5)
energy/Hartrees
-306.9652454 -308.6819625 -308.7389645
-310.786397
-309.6927315 -309.8006391
-
a
b
For atom-numbering scheme, see Figure 1. Figures in parentheses are the estimated standard deviations.
Ta ble 5. Th eor etica l (MP 2/6-311G* level) Str u ctu r a l P a r a m eter s (r /p m , /°) for Con for m er s of [4]Den d r a len ea
conformer (1)
(C2) a,a (s)
conformer (2)
(C2) s,s (a)
conformer (3)
conformer (4)
conformer (5)
(C2) s,s (s)
parameterb
C(2)sC(3)
C(3)sC(4)
C(1)dC(2)
(C1)c a,s (s)
(C1)c s,a (a)
147.0
149.2
134.5
135.1
108.6
121.3
124.6
119.3
-173.4
-173.4
118.9
120.1
74.0
147.9
148.2
134.4
135.3
108.6
121.2
123.5
119.6
-48.5
-48.5
118.0
121.2
158.4
147.5
149.1
134.5
135.0
108.6
121.2
124.1
119.4
-168.3
-39.7
117.3
120.7
69.5
147.6
149.0
134.5
135.3
108.6
121.2
125.1
119.1
-45.2
164.5
118.8
120.7
137.8
147.9
149.1
134.4
134.9
108.6
121.1
123.4
118.8
45.6
C(3)dC(7)
CsH (mean)
CdCH(H) (mean)
C(3)C(2)C(1)
H(2)C(2)C(1)
C(7)C(3)C(2)C(1)
C(8)C(4)C(5)C(6)
C(4)C(3)C(2)
C(4)C(3)C(7)
C(2)C(3)C(4)C(5)
45.6
117.9
120.6
-56.4
energy/Hartrees
rel energyd/kJ mol-1
% at 293 K
-309.8006391
0.00
-309.7975954
8.63
-309.7974499
8.37
-309.7969414
10.27
-309.7921958
21.47
89.05
2.58
5.73
2.63
0.01
a
b
d
Key: a ) anti, s ) syn. For atom-numbering scheme, see Figure 1. c Mean value as for C2 symmetry. Corrected for zero-point
energy (SCF/6-31G* scaled by 0.9).
generally varying over a range of less than 1 pm with
different basis sets. For example, C(3)dC(7) changes
from 132.3 to 132.6 to 132.5 pm for the SCF calculations
employing the 3-21G*, 6-31G*, and 6-311G* basis sets,
respectively. The introduction of electron correlation
leads to the expected lengthening of the CdC bonds by
ca. 2 pm.23 Bond and dihedral angles were consistent to
1-2°, except for the C(2)C(3)C(4)C(5) dihedral angle
which was reduced from ca. 79° to ca. 74° when the effects
of electron correlation were taken into consideration at
the MP2 level.
same general trends as observed for conformer (1), and
are not detailed here. Structural parameters for con-
formers (2-5) at the MP2/6-311G* level are presented
in Table 5.
GED Refin em en t. The rR° structure of [4]dendralene
was refined. For each conformer (1-5), a harmonic
vibrational force field was computed at the HF/6-31G*
level, and the Cartesian force constants were transformed
into those described by a set of symmetry coordinates
using the program ASYM40.24 As a full analysis of
experimental vibrational frequencies is not available for
the compound, it was not possible to scale the theoretical
force constants on this basis. Instead, as the best
Variations in the geometrical parameters of each of the
four conformers (2-5) with theoretical level exhibit the
(23) Hehre, W. H.; Radom, L.; Schleyer P. v. R.; Pople, J . A. Ab Initio
Molecular Orbital Theory; J ohn Wiley & Sons, Inc: New York, 1986.
(24) ASYM40 is an updated version of ASYM20: Hedberg, L.; Mills,
I. M. J . Mol. Spectrosc. 1993, 160, 117.

Gas-Phase Structure of [4]Dendralene
J . Org. Chem., Vol. 62, No. 9, 1997 2771
Ta ble 6. In ter a tom ic Dista n ces (r a /p m ) a n d Am p litu d es
of Vibr a tion (u /p m )
alternative, empirical scale factors were employed, based
on the work of Pulay et al. for butadiene:25 0.87 for CsH
and CdC bond stretches, 0.92 for CsC bond stretches,
and 0.80 for angle bends and torsions. Values for the
root-mean-square amplitudes of vibration (u) and per-
pendicular amplitude corrections (K) were then derived
from the scaled force constants using ASYM40.24
Initial refinements of the electron-diffraction pattern
assumed the presence of a single conformer with C2
symmetry, for which starting values of geometrical
parameters were taken from the structure of conformer
(1) optimized ab initio at the MP2/6-31G* level. It was
possible to refine nine of the twelve geometrical param-
eters (Table 2); introduction of parameters defining
differences between the CsC and CdC bond lengths, i.e.
p3 and p4, and of p8, the angle H(2)C(1)C(2), either caused
the refinement to become unstable (large oscillations in
the RG factor between cycles) or led such parameters to
adopt unrealistic values. These parameters were refined
subsequently using flexible restraints.
Flexible restraints may allow the refinement of pa-
rameters which would otherwise have to be fixed.26
Estimates of the values of these restrained quantities and
their uncertainties are used as additional observations
in a combined analysis similar to those routinely carried
out for electron-diffraction data combined with rotation
constants and/or dipolar coupling constants.27 The values
and uncertainties for the extra observations are derived
from another method such as X-ray diffraction or theo-
retical computations. All geometrical parameters are
then included in the refinements. In cases where a
restrained parameter is also a refinable parameter, if the
intensity pattern contains useful information concerning
the parameter, it will refine with an esd less than the
uncertainty in the corresponding additional observation.
However, if there is essentially no relevant information,
the parameter will refine with an esd equal to the
uncertainty of the extra observation and its refined value
will equal that of the restraint. In this case, the
parameter can simply be fixed, in the knowledge that
doing this does not influence either the magnitudes or
the esds of other parameters. In some cases, because
increasing the number of refining parameters allows all
effects of correlation to be considered, some esds may
increase. Overall, this approach utilizes all available
data as fully as possible and returns more realistic esds
for refining parameters; the unknown effects of correla-
tion with otherwise fixed parameters are revealed and
included.
no.
atom paira,b
distancec
amplitudec
1
2
3
4
5
6
7
8
9
C(2)sC(3)
147.8(2)
5.1(1)
5.2 (tied to u1)
C(3)sC(4)
C(1)dC(2)
C(3)dC(7)
C(7)sH(7)
C(7)sH(7′)
C(2)sH(2)
C(1)sH(1′)
C(1)sH(1)
149.7(3)
134.9(1)
134.6(1)
110.7(1)
110.6(1)
112.6(1)
112.6(1)
110.8(1)
4.4 (1)
}
8.3(1)
}
10 C‚‚‚H (two-bond)
11 C(3)‚‚‚C(8)
12 C(2)‚‚‚C(7)
13 C(1)‚‚‚C(3)
14 C(2)‚‚‚C(4)
15 C(1)‚‚‚C(4)
16 C(7)‚‚‚C(8)
17 C(2)‚‚‚C(5)
18 C(1)‚‚‚C(5)
19 C(2)‚‚‚C(8)
20 C(1)‚‚‚C(6)
21 C(1)‚‚‚C(7)
22 C(1)‚‚‚C(8)
211.0(4)-219.2(19) 10.0-10.5(2)
242.9(8)
248.2(5)
248.7(4)
256.3(7)
296.3(16)
306.9(23)
329.5(22)
339.1(23)
347.2(16)
352.4(48)
370.1(4)
376.2(31)
6.6(4)
6.4
6.3 (tied to u11
)
}
6.9
11.7(13)
11.4 (tied to u15
12.5 (tied to u18
16.6(20)
)
)
12.5(14)
18.2 (tied to u19
7.0(4)
)
21.4(41)
23 C‚‚‚H (three/four-bond) 259.7-272.9
24 C‚‚‚H (three-bond)
25 C‚‚‚H (four-bond)
15.6-21.0(14)d
10.1-10.7(10)d
13.6-15.4(12)d
343.9-352.8
402.7-408.2
a
b
For atom-numbering scheme, see Figure 1. Other C‚‚‚H and
H‚‚‚H nonbonded distances were included in the refinements but
are not listed here. c Values in parentheses are the estimated
d
standard deviations. Refined as a group; one amplitude refining
with others having a fixed ratio relative to it.
A careful study of models defining more than one
conformer was undertaken. For each, the change in the
RG factor was followed as a function of the composition
of the mixture. For all compositions, the RG factor was
found to be greater than that for the single conformer
alone, i.e. no other minima were located on the RG vs %
composition surface generated. At the 95% confidence
level,28 the gas-phase sample of [4]dendralene was found
experimentally to consist of 100 ( 2% of conformer (1).
Values of the principal interatomic distances for the
final refinement (RG ) 0.019, RD ) 0.021) are listed in
Table 6 and the most significant values of the least-
squares correlation matrix are given in Table 7. The
experimental and difference radial-distribution curves
are shown in Figure 2 and the molecular-scattering
intensities in Figure 3. Cartesian coordinates are in-
cluded as part of the Supporting Information.
Discu ssion
In the gas phase at 293 K, [4]dendralene is found
experimentally by electron diffraction to consist of a
single conformer with C2 symmetry (Figure 1a). The
butadiene groups, which are twisted by 5.2(28)° from a
perfectly planar conformation, have anti arrangements
of the CdC bonds. They are orientated at 71.9(19)° to
one another relative to the anti,anti (syn) starting posi-
tion in Scheme 2.
The principal structural parameters predicted by the
ab initio computations are given in Table 4. At the MP2/
6-311G* level, the values defining the equilibrium ge-
ometry (re) for conformer (1) are in very good agreement
with those refined from the electron-diffraction pattern
(rR°). Four other minima, apart from that observed
experimentally, were located on the PES; relative ener-
gies at a range of theoretical levels are shown in Table
Using flexible restraints, it was possible to refine p3,
p4, and p8. These were restrained directly with values
p3 ) 2.3 ( 0.5 pm, p4 ) 0.50 ( 0.25 pm, and p8 ) 119.3
( 2.0°.
In addition, eighteen amplitudes of vibration were
included in the final refinements, of which seven (C‚‚‚H
distances > 300 pm) were refined using restraints. For
amplitudes tied together, the ratio of amplitudes within
each group was fixed at the theoretical value.
(25) Pulay, P.; Fogarasi, G.; Pongor, G.; Boggs, J . E.; Vargha, A. J .
Am. Chem. Soc. 1983, 105, 7037.
(26) (a) Blake, A. J .; Brain, P. T.; McNab, H.; Miller, J .; Morrison,
C. A.; Parsons, S.; Rankin, D. W. H.; Robertson, H. E.; Smart, B. A. J .
Phys. Chem. 1996, 100, 12280. (b) Brain, P. T.; Morrison, C. A.;
Parsons, S.; Rankin, D. W. H. J . Chem. Soc., Dalton Trans. 1996, 4589.
(27) For example, see: Abdo, B. T.; Alberts, I. L.; Attfield, C. J .;
Banks, R. E.; Blake, A. J .; Brain, P. T.; Cox, A. P.; Pulham, C. R.;
Rankin, D. W. H.; Robertson, R. E.; Murtagh, V.; Heppeler, A.;
Morrison, C. J . Am. Chem. Soc. 1996, 118, 209.
(28) Hamilton, W. C. Acta Crystallogr. 1965, 18, 502.

2772 J . Org. Chem., Vol. 62, No. 9, 1997
Brain et al.
Ta ble 7. Lea st-squ a r es Cor r ela tion Ma tr ix (× 100)a
consistent with these conformers not being detected
experimentally by electron diffraction.
The details of the structures of conformers (2-5)
optimized at the MP2/6-311G* level are given in Table 5
and illustrated in Figure 1. Conformers (1) and (3) and
conformers (2) and (4) may be viewed as pairs in that
each constitutes a reduction in symmetry from C2 to C1
brought about by changing one butadiene group from anti
to syn, or vice versa, while retaining the relative orienta-
tion of the butadienes, e.g. for conformers (1) and (3), a,
a (s) f a, s (s). Similarly, conformers (5) and (2) are
related in that they have syn conformations for the
butadienes but different relative orientations of these
groups.
With the exception of conformer (1), all the conformers
show large deviations from perfect anti and, more
particularly, perfect syn conformations of one or both of
the butadienyl groups. For example, in conformer (2) the
deviation is 48.5° from syn [C(7)C(3)C(2)C(1)], and in
conformer(3) the deviation is 11.7° from anti [C(7)C(3)C-
(2)C(1)] and 39.7° from syn [C(8)C(4)C(5)C(6)]. The level
of conjugation within each butadienyl group is thus
variable, decreasing with increased distortion of the
conformation. For conformers (2-5) the distortions are
greater than for conformer (1) and all have slightly longer
C(2)-C(3) distances, as expected.
Notwithstanding the dihedral angles, the bond lengths
and angles describing the five conformers are very
similar, indicative of steric factors (C‚‚‚C, C‚‚‚H, and
H‚‚‚H nonbonded interactions) being responsible pre-
dominantly for the disposition of the groups in each.
Cross conjugation of the butadiene groups would be
evidenced by a shortening of C(3)-C(4) relative to C(2)-
C(3) and expected only for dihedral angles C(2)C(3)C(4)C-
(5) approaching 0° or 180°. Of the five conformers,
conformer (2) offers the best approximation to this
criterion, with C(2)C(3)C(4)C(5) ) 158.4°, and conformer
(1) the worst, with C(2)C(3)C(4)C(5) ) 74°. Conformer
(2) demonstrates the smallest C-C difference (0.3 pm)
[it has the shortest C(3)-C(4) bond length (148.2 pm)],
and conformer (1) the largest (2.2 pm) [it has the longest
C(3)-C(4) bond length (149.2 pm)], consistent with a
π-bonding contribution to C(3)-C(4) in conformer (2). The
UV spectrum of [4]dendralene, whether consisting pre-
dominantly (ab initio) or completely (GED) of conformer
(1), would thus be expected to be similar to that of
butadiene, as is found experimentally (λmax ) 216.5 nm;
cf. butadiene λmax ) 217 nm).5,7
F igu r e 2. Observed and final weighted difference radial-
distribution curves for [4]dendralene, conformer (1). Before
Fourier inversion the data were multiplied by s exp[(-
0.00002s2)/(ZC - fC)(ZC - fC)].
[4]Dendralene may be considered as a 2-substituted
anti butadiene (Table 8); structural parameters of related
compounds are given in Table 8. On replacement of H
at C(2) in butadiene,29 the CdCsC ipso angle narrows
and the other CsCdC angle widens, except for X )
C(dCH2)CHdCH2, [4]dendralene, where it is unchanged.
This is consistent with increased steric repulsions relative
to butadiene on substitution; such interactions are pre-
sumably less marked for [4]dendralene where the near
planar butadiene groups lie almost orthogonal to one
another [C(2)C(3)C(4)C(5) ) 71.7(19)°]. The attenuation
of the C-C bond on substitution with particularly bulky
F igu r e 3. Observed and final weighted difference molecular-
scattering intensity curves for [4]dendralene, conformer (1).
Nozzle-to-plate distances were (a) 285, and (b) 128 mm.
t
groups, i.e. CHdCH2,4 C(dCH2)CHdCH2 and Bu,31 lends
further support to the premise that nonbonded repul-
3. At the MP2/6-311G* level, none of these additional
conformers is predicted to constitute more than 6% of the
gas-phase mixture. Such small proportions, coupled with
the very similar interatomic-distance distribution, are
(29) Kveseth, K.; Seip, R.; Kohl, D. Acta Chem. Scand. 1980, A34,
31.

Gas-Phase Structure of [4]Dendralene
J . Org. Chem., Vol. 62, No. 9, 1997 2773
Ta ble 8. GED Str u ctu r a l P a r a m eter s (r a ) for Som e 2-Su bstitu ted An ti Bu ta d ien esa
X
r(CdC)b
r(C-C)
CdCsC (ipso)
CsCdC
ref
H
CH3
CHdCH2
134.9(1)
134.0(1)
134.4(19)
134.8(1)
134.5(3)
134.2(2)
146.7(2)
146.3(2)
147.9(3)
148.4(2)
148.5(10)
146.9(3)
124.4(1)
121.4(3)
118.7(27)
123.2(4)
121.7(12)
123.5(1)
124.4(1)
127.3(3)
126.6(31)
124.4(3)
126.2(12)
125.6(2)
29
30
4
C(dCH2)CHdCH2
this work
31
32
tBu
Clc
a
b
Values in parentheses are the estimated standard deviations. Where two different CdC distances have been refined, a mean value
is shown. c Combined GED/microwave analysis.
sions, rather than electronic factors, account primarily
for the conformations of these compounds.31,33
and (iii) the award of a postdoctoral fellowship to B.A.S.
and funds for the purchase of a workstation (grant GR/
K04194). We thank Dr. L. Hedberg (Oregon State
University) for a copy of the program ASYM40 and Dr.
M. H. Palmer (University of Edinburgh) for performing
preliminary computations.
Ack n ow led gm en t. We thank the EPSRC for (i)
financial support of the Edinburgh Electron Diffraction
Service (grant GR/K44411) including research fellow-
ships to P.T.B. and H.E.R., (ii) provision of the mi-
crodensitometer facilities at the Daresbury Laboratory,
Su p p or tin g In for m a tion Ava ila ble: Listings of (a) ab-
solute energies for the theoretical computations and (b) atomic
coordinates for the GED and theoretically optimized geom-
etries (7 pages). This material is contained in libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
(30) Trætteberg, M.; Paulen, G.; Cyvin, S. J .; Panchenko, Y. N.;
Mochalov, V. I. J . Mol. Struct. 1984, 116, 141.
(31) Trætteberg, M.; Hopf, H.; Ha¨nel, R. Chem. Ber. 1994, 127, 1459.
(32) Gundersen, G.; Thomassen, H. G.; Boggs, J . E.; Peng, C. J . Mol.
Struct. 1991, 243, 385.
(33) Trætteberg, M.; Bakken, P.; Hopf, H.; Ha¨nel, R. Chem. Ber.
1994, 127, 1469.
J O962091H