UV photoelectron spectroscopic and computational study of 7-substituted cycloheptatrienes

Article abstract of DOI:10.1139/V08-041

The He(I) photoeiectron (PE) spectra of cycloheptatriene (1), 7-methylcycloheptatriene (2), 7-methoxycycloheptatriene (3), 7-methylthiocycloheptatriene (4), and 7-diemthylaminocycloheptatriene (5) were recorded and interpreted using MO energies and ionization potentials acquired from B3PW91 calculations. Partial simulated PE spectra were in good agreement with the experimental results. The axial and equatorial conformers have distinct PE spectra as illustrated by simulation. The PE spectra of 2, 3, and 5 are representative of the equatorial conformers, while the PE spectrum of 4 was in accord with a 1:1 mixture of equatorial and axial conformational isomers based on spectral simulation. The calculated free energy differences between the equatorial and axial conformers are in the order of 2 kcal mol^-1 (1 cal = 4.184 J) for compounds 2, 3, and 5, where the equatorial conformational isomer is lowest in energy. In the case of 4, the equatorial conformer was only marginally lower in energy than its axial counterpart, where the free energy difference is 0.1 kcal mol^-1.

Full text of DOI:10.1139/V08-041

4
44
UV photoelectron spectroscopic and
computational study of 7-substituted
cycloheptatrienes
T. Bajorek and N. H. Werstiuk
Abstract: The He(I) photoelectron (PE) spectra of cycloheptatriene (1), 7-methylcycloheptatriene (2), 7-
methoxycycloheptatriene (3), 7-methylthiocycloheptatriene (4), and 7-diemthylaminocycloheptatriene (5) were recorded
and interpreted using MO energies and ionization potentials acquired from B3PW91 calculations. Partial simulated PE
spectra were in good agreement with the experimental results. The axial and equatorial conformers have distinct PE
spectra as illustrated by simulation. The PE spectra of 2, 3, and 5 are representative of the equatorial conformers,
while the PE spectrum of 4 was in accord with a 1:1 mixture of equatorial and axial conformational isomers based on
spectral simulation. The calculated free energy differences between the equatorial and axial conformers are in the order
of 2 kcal mol–1 (1 cal = 4.184 J) for compounds 2, 3, and 5, where the equatorial conformational isomer is lowest in
energy. In the case of 4, the equatorial conformer was only marginally lower in energy than its axial counterpart,
–
1
where the free energy difference is 0.1 kcal mol .
Key words: 7-substituted cycloheptatrienes, He(I) photoelectron spectroscopy, B3PW91, spectral simulation.
Résumé : On a enregistré les spectres photoélectroniques (PE) He(I) du cycloheptatriène (1), du 7-méthyl-
cycloheptatriène (2), du 7-méthoxycycloheptatriène (3) et du 7-méthylthiocycloheptatriène (4) et on les interprète en
faisant appel aux énergies des orbitales moléculaires (OM) et des potentiels d’ionisation obtenus par des calculs
B3PW91. Les spectres PE simulés partiels sont en bon accord avec les résultats expérimentaux. Tels qu’illustrés par la
simulation, les conformères axiaux et équatoriaux présentent des spectres PE distincts. Les spectres PE des produits 2,
3
et 5 sont représentatifs de conformères équatoriaux alors que le spectre PE du composé 4 correspond au spectre de
simulation d’un mélange 1:1 des isomères conformationnels équatoriaux et axiaux. Les différences d’énergie libre entre
–
1
les conformères équatorial et axial sont de l’ordre de 2 kcal mol pour les conformères équatoriaux des composés 2, 3
et 5. Dans le cas du produit 4, le conformère équatorial n’est que marginalement favorisé par une différence d’énergie
de 0,1 kcal mol–1 par rapport au conformère axial.
Mots-clés : cycloheptatriènes substitués en position 7, spectroscopie photoélectronique He(I), B3PW91, simulation spectrale.
[
Traduit par la Rédaction] Bajorek and Werstiuk 450
Introduction
Fig. 1. Schematic representation of equatorial and axial isomers
of cycloheptatrienes.
Cycloheptatrienes substituted at the 7-position may exist
in two conformations, the axial and the equatorial
conformational isomers, illustrated in Fig. 1. 7-
Silylcycloheptatriene and 7-methylcycloheptatriene (2) have
previously been under computational investigation, where it
has been reported that the axial conformers were predicted
to be more stable than the equatorial conformers by nearly
–
1
2
kcal mol (1) (1 cal = 4.184 J). This was contrary to our
calculations that showed that the difference in energy be-
tween axial and equatorial 7-silylcycloheptatriene is nearly
nil. In the case of 7-methylcycloheptatriene (2), we found
the equatorial conformational isomer to be lower in energy
than the axial conformer. The preference of the axial confor-
mation in a 7-substituted cycloheptatriene has previously
been documented for triphenyl-7-cyclohepta-1,3,5-trienyltin,
as evidenced by its crystal structure (2, 3). Given our long-
standing interest in studying conformational properties in the
gas phase with ultraviolet (UV) photoelectron (PE) spectroscopy
Received 19 December 2007. Accepted 13 February 2008.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 11 April 2008.
(4) and our work on identifying E,Z isomers of cyclohepta-2,4-
This paper is dedicated to the memory of Keith Yates.
dien-1-one (5) and N,N-dimethylcyclohepta-1,3,5-trien-1-amine
(6), we initiated a computational and UV PE spectroscopic study
of cycloheptatriene (1), 7-methylcycloheptatriene (2), 7-
methoxycycloheptatriene (3), 7-methylthiocycloheptatriene (4),
and 7-dimethylaminocycloheptatriene (5) (Fig. 2). We under-
1
T. Bajorek and N.H. Werstiuk. Department of Chemistry,
McMaster University, 1280 Main St. West, Hamilton ON
L8S 4M1, Canada.
1
Corresponding author (werstiuk@mcmaster.ca).
Can. J. Chem. 86: 444–450 (2008)
doi:10.1139/V08-041
© 2008 NRC Canada

Bajorek and Werstiuk
445
Fig. 2. 7-Substituted cycloheptatrienes.
Calculated conformational properties
The equatorial conformer is the most stable conformational
isomer of cycloheptatrienes 2, 3, and 5; the substituent is situ-
ated in the plane of the 7-membered ring rather than perpen-
dicular to the ring as in the axial conformation, as shown in
Fig. 1. For the 7-methylthiocycloheptatriene (4), the axial
conformer is only marginally lower in energy than its equato-
rial counterpart. Of course, the “equatorial” and “axial” con-
formations are identical in the case of cycloheptatriene (1).
Our B3PW91/6–311+G(2d,p) calculations place the 2-ax-
took the study to determine whether it would be possible to
distinguish between the axial and equatorial conformers in
the gas phase using PE spectroscopy coupled with DFT cal-
culations and thereby obtain information on the
conformational preferences for the compounds, while docu-
menting the PE spectra of 2–5 that had not yet been pub-
lished. In this paper we present and discuss the results of
this study.
–
1
ial conformer 1.9 kcal mol higher in energy — based on
Eelec — than the equatorial conformer, which is contrary to
∆
previously reported findings (1). Donovan and White have
carried out semi-empirical, ab initio, and DFT calculations
on 7-methoxycycloheptatriene (3), finding that the equato-
–
1
rial conformation is favored by 1.2 kcal mol at BLYP/6–
1G(d) (14). Our calculations agree with those of Donovan
and White and predict the equatorial conformer of 3 to be
3
–
1
1
.8 kcal mol lower in energy than the axial conformer at
the B3PW91/6–311+G(2d,p) level. The difference in energy
between the equatorial and axial conformational isomers of
Results
–
1
4
is only 0.8 kcal mol , with the axial conformer only
The experimental UV PE spectra of 1–5 as well as the
corresponding partial simulated spectra of their axial and
equatorial conformers are shown in Fig. 3. Table 1 lists the
B3PW91/6–311+G(2d,p) total energies of 1, the axial and
equatorial conformational isomers of 2–5, the corresponding
radical cations, and the derived first ionization potentials.
slightly more stable than the equatorial one. If the ZPVE
correction is taken into account, then this difference de-
–
1
creases to 0.1 kcal mol . Based on free energies, the equa-
torial conformer is marginally lower in energy (–0.2 kcal
–
1
mol ) than the axial conformer. The equatorial conformer
–
1
of 5 is 1.9 kcal mol lower in energy than the axial con-
former.
Although other methods
TDDFT) (7, 8) and outer valence Green’s function (OVGF)
9, 10) — are available for predicting IPs, we preferred to
—
time-dependent DFT
Cycloheptatrienes
2–5
undergo
axial–equatorial
(
(
conformational isomerism. The B3PW91 ZPVE corrected
total energies and relative energies (∆E and ∆G298) of the
use the method developed in these laboratories (11). In our
approach, the first vertical IP, calculated as the difference in
energy between ground state and the radical cation fixed in
the ground-state optimized geometry, is used to anchor the
Kohn–Sham orbital energies. Our approach is also useful for
obtaining an estimate of the first-band adiabatic IP simply
by optimizing the radical cation. Using our method to obtain
the IPs (11) and our program PESPEC (12a, 12b), we have
successfully simulated the PE spectra of a variety of fairly
large molecules, some of which exhibit conformational mo-
bility. In our approach, we compare the accuracy of the cal-
culations and the “shapes” of the predicted and experimental
spectra. Our method has also been shown to perform very
well by other investigators (13). Consequently, we saw no
need to do an extensive time-consuming study with all the
other computational methods.
0
conformers are shown in Table 3. Cycloheptarienes 2, 3, and
5
are predicted to have the equatorial conformer lowest in
energy; in the case of 4, the axial conformation is margin-
ally lower in energy. Energy barriers for the ring flip from
equatorial to axial are collected in Table 3.
The B3PW91 potential energy surface for the equatorial
cycloheptatrienes 3 and 4, obtained by varying the dihedral
angle between H-7 and the Me group of –OMe and –SMe,
revealed that the gauche conformation is lowest in energy in
the three cycloheptatrienes, as shown in Fig. 4. It is also the
lowest energy conformation in the axial conformational iso-
mers with the exception of 4-axial where the anti conforma-
tion is favored.
Discussion
Table 2 lists the first five calculated (B3PW91/6–
3
11+G(2d,p)) and first three experimental vertical IPs of com-
The PE spectrum of 1, shown in Fig. 3a, exhibits three
distinct ionization potential bands at 8.6, 9.5, and 10.9 eV;
the calculated IPs and consequently the partial simulated PE
spectrum displayed as Fig. 3b are in good agreement with
the experimental spectrum.
pounds 1–5. Overall the differences between the calculated
and experimental IPs is <0.4 eV. Given the closeness of the
first two IPs, for the case of 3 and 4 we used the average val-
ues of the first two IPs given in parentheses in Table 2 for the
comparison. We also calculated the first five IPs with OVGF.
While the IPs moved to slightly higher values (Table 1S of
the Supplementary Data), there are no significant differences
The PE spectra of 2, 3, and 5 are representative of mostly
the equatorial conformers of these compounds. Based on our
B3PW91 calculations, the equatorial conformers are on the
2
–1
in the relative separations of the DFT and OVGF IPs.
order of 2 kcal mol lower in free energy than their axial
2
Supplementary data for this article are available on the journal Web site (canjchem.nrc.ca) or may be purchased from the Depository of Un-
published Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 3739. For more in-
formation on obtaining material refer to cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml.
©
2008 NRC Canada

4
46
Can. J. Chem. Vol. 86, 2008
Fig. 3. Experimental PE spectra: (a) 1, (c) 2, (g) 3, (k) 4, (o) 5; partial simulated (B3PW91) PE spectra: (b) 1, (d) 2 equatorial, (e) 2
axial, (f) 2 96% equatorial – 4% axial, (h) 3 equatorial, (i) 3 axial, (j) 3 96% equatorial – 4% axial, (l) 4 equatorial, (m) 4 axial, (n)
4
57% equatorial – 43% axial, (p) 5 equatorial, (q) 5 axial, (r) 5 96% equatorial – 4% axial.
Table 1. Total energies Eelec (Hartrees) of equatorial and axial
The partial simulated PE spectra of the equatorial and ax-
ial conformational isomers as well as the combination of
96% equatorial conformer and 4% axial conformer of 3 are
displayed in Figs. 3h, 3i, and 3j, respectively. Although there
is a difference in the relative intensities of first bands of the
simulated and experimental spectra, the general features of
the simulated spectrum 3j and the experimental spectrum 3g
correlate reasonably well.
cycloheptatrienes 1, 2, 3, 4, and 5 (N electrons), their radical
a
cations (N-1 electrons), and calculated first IP (eV).
v
Eelec(N) (Hartree)
Eelec(N-1)a (Hartree)
IP (eV)
v
1
2
2
3
3
4
4
5
5
–271.481 326
–310.793 434
–310.790 389
–385.989 541
–385.986 682
–708.954 913
–708.956 163
–405.428 979
–405.425 890
–271.178 032
–310.491 535
–310.493 891
–385.682 710
–385.680 381
–708.655 627
–708.669 426
–405.133 707
–405.142 044
8.25
8.21
8.07
8.35
8.33
8.14
7.80
8.03
7.72
eb
a
c
e
a
e
a
e
a
a
The
experimental
spectrum
of
7-
methylthiocycloheptatriene is displayed as Fig. 3k. Fig-
ures 3l and 3m are the simulated spectra of the equatorial
and axial conformers of 4, respectively. According to our
calculations, the equatorial conformer is only marginally
–
1
(∆G = 0.16 kcal mol ) lower in energy than the axial iso-
2
98
mer so the equatorial conformer should be slightly preferred
(55%–60%) at equilibrium (15). In fact, the simulated spec-
trum 3n (57% equatorial + 43% axial) correlates better with
the experimental spectrum than spectra 3l and 3m.
Single point energy calculations on the optimized molecule (N elec-
trons) geometries with an electron removed.
b
e = equatorial.
c
a = axial.
As far as the PE spectra of 5 are concerned, the simulated
spectra of equatorial 5 (Fig. 3p) and a 94% equatorial – 4%
axial 5 (Fig. 3r) correlate superbly with the experimental
spectrum displayed in Fig. 3o.
To confirm our analysis based on the relative energies of
single equatorial and axial conformers, we computed the
barriers for ring flipping to obtain information on the nature
of the ring-flipping potential energy surface. This is impor-
tant because the relative intensities of PE bands can depend
on the conformer populations and on the variation in their
counterparts so the equatorial conformations are expected to
be preferentially populated at equilibrium. The partial simu-
lated PE spectra of the equatorial and axial conformational iso-
mers as well as the combination of 96% equatorial conformer
and 4% axial conformer — these values are based on the mag-
–1
nitude of the equilibrium constant with ∆G = 2.0 kcal mol
(15) — of 2 are displayed as Figs. 3d, 3e, and 3f, respec-
tively. Both partial spectra 3d and 3f (96% equatorial + 4%
axial) correlate well with the experimental spectrum 3c.
©
2008 NRC Canada

Bajorek and Werstiuk
Table 2. Experimental and calculated (B3PW91/6–311G(2d,p)) vertical IPs of compounds 1–5.
447
IPv
1
2
B3PW91
Experimental
B3PW91
8.25
8.6
8.07
8.21
8.5
8.33
8.35
8.6
7.80
8.14
8.7
9.15
9.5
9.17
9.01
9.4
8.98
9.03
9.6
8.08
8.44
10.74
10.9
10.58
10.55
10.8
9.28
9.43
10.0
9.75
9.21
9.7
11.46
11.76
axial
equatorial
10.93
11.28
11.32
11.49
Experimental
B3PW91
3
4
5
axial
equatorial
10.53
10.71
11.1
10.51
10.55
11.1
10.79
10.68
10.9
11.08
11.15
Experimental
B3PW91
a
a
axial
equatorial
7.94
8.29
11.18
10.89
Experimental
B3PW91
a
a
axial
equatorial
7.72
8.03
8.6
8.26
8.29
7.99
8.16
9.02
8.85
9.4
11.10
11.33
Experimental
a
Average of the first and second IPs.
Table 3. Thermochemical parameters for equatorial and axial cycloheptatriene conformers of 2–5 and ring-flip transition
states at B3PW91/6–311+G(2d,p) (1 cal = 4.184 J).
Eeleca (Hartrees) ∆Eelec (kcal mol–1) E0b (Hartrees) ∆E (kcal mol )
–1
G298c (Hartrees) ∆G298 (kcal mol–1)
0
2
ed
TS2
a
e
–310.793 434
–310.782 074
–310.790 389
–385.989 541
–385.979 864
–385.986 682
–708.954 913
–708.947 945
–708.956 163
–405.428 979
N/A
0.00
7.13
1.91
0.00
6.07
1.79
0.78
5.16
0.00
0.00
N/A
1.94
–310.638 097
–310.626 464
–310.634 563
–385.829 852
–385.819 853
–385.826 454
–708.798 947
–708.791 693
–708.799 099
–405.228 892
N/A
0.00
7.30
2.22
0.00
6.27
2.13
0.10
4.65
0.00
0.00
N/A
2.11
–310.668 974
–310.656 828
–310.665 445
–385.862 704
–385.852 353
–385.859 405
–708.833 290
–708.825 011
–708.833 039
–405.262 984
N/A
0.00
7.62
2.21
0.00
6.50
2.07
0.00
5.20
0.16
0.00
N/A
2.07
e
2
3
TS3
3
4
TS4
4
5
TS5
a
e
a
e
f
5
a
–405.425 890
–405.225 525
–405.259 681
a
Eelec is the uncorrected total energy in Hartrees.
E = E + ZPE.
b
c
0
elec
G = H – TS; H = E + RT; E = E + E + Erot + Etrans
.
0
vib
d
e = equatorial.
e
a = axial.
f
Transition state could not be located.
IPs (4d, 12a, 12b, 16). The equatorial–axial ring-flipping
barriers for 2 and 3 are 7.62 and 6.50 kcal mol , respec-
significantly populated at ambient temperature and contrib-
ute to the PE spectrum.
–
1
tively, (Table 3). For 4-equatorial to 4-axial, it is 5.20 kcal
It should be pointed out that the partial simulated PE
spectra of 3 and 4 have been obtained for the lowest energy
rotomers. Figures 4a–4d show the potential energy surfaces
for the rotation of the –OMe and –SMe substituents; the
gauche conformations are lowest in energy in the cases of 3-
equatorial, 3-axial, and 4-equatorial. 4-Axial is the lowest-
energy conformer in the case of the anti orientation. Plots of
the orbital energy versus dihedral angle for the five MOs,
HOMO–HOMO-4 of 3-equatorial and 4-equatorial, show
small energy differences in orbital energies when the
substituent is rotated (Fig. 5a). The simulation of partial PE
spectra based on this angular dependence data and the poten-
tial energy surfaces displayed in Figs. 4a and 4b, using our
program PESPEC (12a), produced simulated spectra that did
not differ significantly — not unexpected given the relative
invariance of the MO energy with torsional angle — from
–
1
mol . A discrete transition state for the ring flip of 5 could
not be located, presumably because of the complexity — in-
version at nitrogen is possible — of the potential energy sur-
face. However, after an extensive study we were able to
–
1
locate a “minimum-energy” geometry 7.64 kcal mol higher
in energy than 5, exhibiting a geometry that was close to the
TS geometries of 2, 3, and 4 with respect to the planarity of
the 7-membered ring. The computed low barriers for the
ring flip support the fact that 1–5 are fluxional molecules
(
1–3), and that the equilibrium populations of the equatorial
and axial conformations at ambient temperature are deter-
mined by ∆G. This validates our analysis based on using the
equatorial and axial conformers; given the nature of the po-
tential energy surfaces, only the equatorial and axial con-
formers and those very closely related to them are
©
2008 NRC Canada

4
48
Can. J. Chem. Vol. 86, 2008
Fig. 4. The B3PW91/6–311+G(2d,p) potential energy surface for the rotation of the –OMe (a and b) and –SMe (c and d) substituents
in the equatorial and axial isomers of 3 and 4 (1 cal = 4.184 J).
those presented in Fig. 3 and are not reported in this paper.
Given the nature of the angular dependences of the MO en-
ergies and the torsional potential energy surfaces, the same
situation will hold for 4.
have shown that the experimental PE spectra of compounds
2, 3, and 5 are those of mainly the equatorial conformers
and are in good agreement with experiment. The simulation
has also shown that the PE spectrum of 4 is best represented
by a nearly 1:1 mixture of the equatorial and axial
conformational isomers, which is supported by our
thermochemical calculations on these compounds.
As far as other 7-substituted cycloheptatrienes are con-
cerned, for 7-SiH C H it has been reported that the axial
3
7
7
conformation is more stable than its equatorial counterpart
–
1
by nearly 2 kcal mol , based on computational studies, and
the C–Si bond lengths of the two conformers differ (1). Our
B3PW91/6–311+G(2d,p) calculations indicate that the axial
conformational isomer is only marginally more stable than
Experimental detail
Computational studies
–
1
the equatorial conformer with ∆G2 being 0.05 kcal mol .
98
B3PW91 (17) calculations were performed with Gaussian
3 (18). Calculations on optimized geometries and radical
The C–Si bond lengths are slightly different; for axial the
C–S distance is 1.91 Å and for the equatorial conformer it is
0
cations were carried out using the TIGHT convergence crite-
ria at B3PW91/6–311+G(2d,p). Two-electron integrals and
their derivatives were computed using the ultrafine grid. As
it is well established that the correct geometry optimizations
of molecules containing chalcogen atoms require the addi-
tion of d-polarization functions (19–23), we have used the
above triple zeta split valence basis set augmented by two d-
polarization functions, one p-polarization function on hydro-
gen atoms, as well as one diffuse function on hydrogen at-
oms. The interpretation of the PE spectra was based on our
successfully applied method (11, 24, 25, 26) that is basically
an extension of Koopman’s theorem. The simulated PE spec-
tra of all compounds were obtained by anchoring the Kohn–
Sham MO orbital energies to the calculated first IP using our
Fortran program PESPEC (12(a)). Band intensities are nor-
malized in simulations involving single structures and (or)
conformers. However, in cases where conformations exhibit
1
.89 Å. The axial conformation in a 7-substituted
cycloeheptatriene has been experimentally observed in
triphenyl-7-cyclohepta-1,3,5-trienyltin (2, 3). Its crystal
structure clearly showed the quasi-axial orientation of the
triphenyltin substituent at position 7 of the cycloheptatriene
ring with a carbon–tin bond length of 2.21 Å (2). The trend
towards axial orientations of the polarizable C–Si and C–Sn
bonds suggests that a hyperconjugation interaction plays a
role in determining the axial–equatorial preference.
Conclusions
The photoelectron spectra of cycloheptatrienes 1–5 were
recorded and interpreted with the use of calculations at the
B3PW91 level of theory. Simulated PE spectra of the axial
and equatorial conformational isomers of all compounds
©
2008 NRC Canada

Bajorek and Werstiuk
449
Fig. 5. Orbital energy dependence on dihedral angle for the rotation of the substituent in 3-equatorial and 4-equatorial.
vastly different orbital energies, analyses based on
conformational potential energy surfaces yield simulated
spectra that exhibit intensity variations in the bands due to
the inclusion of conformer populations. Gaussian line shape
with a full-width-at-half-height (FWHH) of 0.4 and a tem-
perature set to 300 K was used to simulate the partial PE
spectra with ten MO energies, HOMO to HOMO-9.
Vertical IPs are accurate to ±0.05 eV. Experimental PE
spectra are displayed in Fig. 3.
Synthesis of cycloheptatrienes
Tropylium bromide was synthesized according to the pro-
cedure by Doering and Knox (28). Cycloheptatriene (1) was
purchased from Sigma-Aldrich and distilled.
UV photoelectron spectroscopic studies
7-Methylthiocycloheptatriene (4)
To a solution of 2 g (0.01 mol/L) of tropylium bromide in
0 mL of methanol, a solution of 0.7 g (0.01 mole) of
A locally built spectrometer (27) was used to acquire the
photoelectron (PE) spectra. A sample of the cycloheptatriene
was placed in an 8 mm od. A pyrex sample tube was exter-
nally attached to the end of a probe. While all spectra were
recorded at the ambient temperature (~20 °C) of the spec-
trometer housing, the temperature of the external sample
was controlled to achieve a suitable source pressure (3–8 ×
1
sodiumthiomethoxide in 30 mL of methanol was added. The
solution was allowed to stir for 0.5 h. Excess
sodiumthiomethoxide was quenched with 50 mL of water.
The resulting solution was extracted with diethylether (3 ×
3
0 mL), which was dried with magnesium sulfate. Removal
–
3
of solvent produced 0.6 (43%) of 7-
methylthiocycloheptatriene. H NMR (CDCl ) δ: 2.08 (s, 3
H), 3.43 (t, 1 H), 5.47 (m, 2 H), 6.21 (m, 2 H), 6.50 (m, 2
g
1
0
mbar, 1 bar = 100 kPa). The sample temperature for
compounds 3, 4, and 5 was 0 °C; for 1 it was –30 °C and for
–15 °C. To obtain the spectra, the probe was inserted into
1
3
2
1
3
H). C (CDCl ) δ: 13.4, 44.0, 124.6, 126.7, 131.1.
the spectrometer housing so the tip was located 10 mm from
the intersecting photon beam. Twenty scans were averaged
3
to obtain the spectra that were calibrated with N (15.6 eV);
7-Methoxycycloheptatriene (3)
A solution of 2 g (0.01 mol/L) of tropylium bromide in
10 mL of methanol was added to a solution of sodium
2
linearity of the scale was ensured through calibrations with
N and CH I (9.54 and 10.16 eV) prior to the experiments.
2
3
©
2008 NRC Canada

4
50
Can. J. Chem. Vol. 86, 2008
methoxide in methanol (1.5 g Na in 20 mL MeOH). The
mixture was stirred for 45 min and filtered to remove precip-
itated sodium bromide. Removal of the solvent afforded yel-
low oil, which was washed with water and extracted with
8. R.E. Stratmann and G.E. Scuseria. J. Chem. Phys. 109. 8218
(1998).
9
.
W. Von Niessen, J. Schirmer, and L.S. Cederbaum. Comput.
Phys. Rep. 1, 57 (1984).
1
0. J.V. Ortiz. J. Chem. Phys. 89, 6348 (1988).
diethylether (3 × 10 mL) and dried with anhydrous MgSO .
4
1
11. H.M. Muchall, N.H. Werstiuk, and B. Choudhury. Can. J.
H NMR (CDCl ) δ: 3.41 (s, 3 H), 3.41 (t, 1 H), 5.51 (m, 2
H), 6.18 (m, 2 H), 6.67 (m, 2 H). C (CDCl ) δ: 56.5, 77.9,
3
1
3
Chem. 76, 221 (1998).
3
1
1
2. (a) N.H. Werstiuk, G. Timmins, J. Ma, and T.A. Wildman. Can.
J. Chem. 70, 1971 (1992); (b) D. Chmielewski, N.H. Werstiuk,
and T.A. Wildman. Can. J. Chem. 71, 1741 (1993).
3. A. Chrostowska, A. Dargelos, V. Lemierre, J-M. Sotiropoulos,
P. Guenot, and J.-C. Guillemin. Angew. Chem. Int. Ed. 43, 873
(2004).
1
23.2, 125.4, 131.1.
7
-Dimethylaminocycloheptatriene (5)
Synthesized according to a procedure by Doering and
1
Knox (17). H NMR (CDCl ) δ: 1.86 (t, 1 H), 2.37 (s, 6 H),
5
4
3
1
3
.45 (m, 2 H), 6.17 (m, 2 H), 6.71 (m, 2 H). C (CDCl ) δ:
3.2, 65.6, 122.8, 123.5, 130.6.
3
14. W.H. Donovan and W.E. White. J. Org. Chem. 61, 969 (1996).
1
5. E.L. Eliel. Stereochemistry of carbon compounds. McGraw-
Hill Book Company Inc., New York. Back cover. 1962.
16. E. Honegger and E. Heilbronner. Chem. Phys. Lett. 81, 615
(1981).
7
-Methylcycloheptatriene (2)
To a vigorously stirred suspension of 2 g. (0.01 mol/L) of
tropylium bromide in dry diethylether, a 1.4 mol/L solution
of methyllithium in diethylether was added dropwize (100%
excess). Exothermic reaction took place upon addition of
methyllithium and reflux was observed. The reaction mix-
ture was allowed to stir for additional 30 min, 20 mL of 2N
HCl was added with care, the mixture was separated and
washed twice with 10 mL portions of 2N NaOH and three
17. A.D. Becke. J. Chem. Phys. 98, 5648 (1993).
18. M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria,
M.A. Robb, J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven,
K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi,
V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian,
J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli,
J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salva-
dor, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich,
A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck,
K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul,
S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith,
M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe,
P.M. W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez,
and J.A. Pople. Gaussian 03 [computer program]. Revision C.02.
Gaussian, Inc., Wallingford, Conn. 2004.
1
5 mL portions of water. The ethereal layer was dried with
anhydrous CaCl . Removal of solvent afforded 0.15 g (14%)
2
1
of 5. H NMR (CDCl ) δ: 1.34 (d, 3 H), 1.64 (m, 1 H), 5.12
3
1
3
(
1
m, 2 H), 6.13 (m, 2 H), 6.63 (m, 2 H). C NMR (CDCl ) δ:
3
9.1, 33.5, 124.4, 128.2, 131.1.
Acknowledgments
We thank the Natural Sciences and Engineering Research
Council of Canada for financial support and gratefully ac-
knowledge the usage of computational resources of
SHARCnet at McMaster University.
1
2
9. A. Hinchliffe. J. Mol. Struct. 55, 127 (1979)
0. W.J. Pietro, M.M. Francl, W.J. Hehre, D.J. DeFrees, J.A. Pople,
and J.S. Binkley. J. Am. Chem. Soc. 104, 5039 (1982).
1. C.J. Marsden and B.J. Smith. J. Mol. Struct. (Theochem.) 105,
References
1
2
.
.
C. Glidewell. J. Organomet. Chem. 266, 25 (1984).
J.E. Weidenborner, R.B. Larrabee, and A.L. Bednowitz. J. Am.
Chem. Soc. 94, 4140 (1972).
2
2
3
85 (1983).
2. A. Yadav, P.R. Surjan, and R.A. Poirier. J. Mol. Struct.
Theochem.) 42, 297 (1988).
3
4
.
.
R.B. Larrabee. J. Organomet. Chem. 74, 313 (1974).
(a) N.H. Werstiuk, J. Ma, M.A. McAllister, T.T. Tidwell, and
D.-C. Zhao. J. Chem. Soc. Faraday Trans. 90, (22), 3383
(
2
2
3. E. Magnusson. J. Am. Chem. Soc. 112, 7940 (1990).
4. H.M. Muchall, N.H. Werstiuk, B. Choudhury, J. Ma,
J. Warkentin, and J.P. Pezacki. Can. J. Chem. 76, 238 (1998).
5. N.H. Werstiuk, H.M.Muchall, J. Ma, and M.T.H. Liu. Can. J.
Chem. 76, 1162 (1998).
(1994); (b) H.M. Muchall, N.H. Werstiuk, J. Ma, T.T. Tidwell,
and K. Sung. Can. J. Chem. 75, 1851 (1997); (c) H.M. Muchall,
N.H. Werstiuk, J.L. Pitters, and M.S. Workentin. Tetrahedron,
2
2
2
2
55, 3757 (1999); (d) H.M. Muchall and N.H. Werstiuk. Can. J.
6. H.M. Muchall and P. Rademacher. J. Mol. Struc. 435, 157
Chem. 84, 1124 (2006).
(
1997).
7. N.H. Werstiuk, D.N. Butler, and E. Shahid. Can. J. Chem. 65,
60 (1986).
8. W.von E. Doering and L.H. Knox. J. Am. Chem. Soc. 79, 352
1957).
5
,
T. Bajorek and N.H. Werstiuk. Chem. Commun. (Cambridge),
6
48 (2002).
7
6
7
.
.
T. Bajorek and N.H. Werstiuk, Can. J. Chem. 82, 80 (2004).
M.E. Casida, C. Jamorski, K.C. Casida, and D.R. Salahub. J.
Chem. Phys. 108, 4439 (1998).
(
©
2008 NRC Canada