been shown to be difficult so far.7 Recently Pascal and co-
workers reported the preparation of 1 in 5% yield by FVP
long periods of time (several months), led to the gradual
disappeareance of the original signals and the emergence of
new peaks in the 1H NMR spectrum. Careful revision of the
spectral data showed that the product, as originally isolated
1
from our reaction, presented an H NMR spectrum (Figure
1b) in which each signal was shifted 0.03-0.09 ppm
of phenanthrene-9,10-dicarboxylic anhydride and its struc-
tural characterization as a D3-symmetric molecular propeller.6
Afterward, we reported a more convenient synthesis of 1
based on the palladium-catalyzed cyclotrimerization of
9,10-didehydrophenanthrene (9,10-phenanthryne, 4) (Scheme
1).8,9
1
Figure 1. (a) H NMR spectra (250 MHz) of 1-C2 in CDCl3 at
298 K; (b) 1-D3 in CDCl3 at 298 K (250 MHz); (c) 1-C2 in 3:1
CS2/CD2Cl2 at 200 K (500 MHz); (d) 1-C2 in 3:1 CS2/CD2Cl2 at
298 K (500 MHz).
downfield with respect to the data reported in the literature,11
while the new emerging peaks were coincidental with these
previously reported data (Figure 1a).
Scheme 1
We attributed this unexpected observation to a conforma-
tional change in 1. Taking into account that the compound
obtained by Pascal had D3 symmetry, as determined by an
X-ray analysis, we reasoned that the product obtained by us
should be the thermodinamically less stable and hitherto
unisolated C2-symmetric conformer. The appearance of only
1
four signals in the H NMR spectrum (Figure 1b,d) could
be due to rapid interconversion of the two possible C2
enantiomeric conformers (I and II) at room temperature. Note
that in this case enantiomerization is achieved by slippage
of two of the outer benzo groups (A and B in I and II, Figure
2) and that in this conversion the C2 axis is rotated by 120°.
Subsequent conversions of this type eventually result in
scrambling of the six benzo groups. Satisfyingly, when we
carried out low-temperature NMR experiments, we observed
at 200 K the splitting of each signal into three signals of
equal intensity (Figure 1c), in accordance with the expected
spectrum for a C2-symmetric conformer. This result was
rather surprising since semiempirical and ab initio calcula-
tions had shown a strong preference (by 5-9 kcal/mol) for
the D3 geometry of 1.6 Intrigued by our finding, we decided
to carry out experimental and computational studies on the
conformational stability of 1. Here we report the results of
these studies.
The kinetics for the conversion of C2 to D3 conformations
(II to III) were determined by 1H NMR in 1,1,2,2-
tetrachloroethane-d2, at constant temperatures 55, 60, 65, 71,
and 82 °C. The rate constants k for each temperature were
determined from the first-order plots shown in Figure 3. From
these data, the Arrhenius plot gave the activation energy Ea
of 22.8 kcal/mol, while the activation parameters ∆Hq (22.0
Further work carried out after our preliminary com-
munication revealed two important new observations: (1)
optimization of the procedure followed for the isolation of
1 from the reaction mixture led to an improvement in the
yield from 39% to a remarkable 68%;10 (2) heating of our
compound in solution, or storage at room temperature for
(7) (a) Hacker, N. P.; McOmie, F. W.; Meunier-Piret, J.; Van Meerssche,
M. J. Chem. Soc., Perkin Trans. 1 1982, 19-23. Other authors reported
the preparation of 1, but no characterization was given: (b) Halleux, A. L.
Br. Patent 852,981, 1960. (c) Barton, J. W.; Grinham, A. R. J. J. Chem.
Soc., Perkin Trans. 1 1972, 634-637.
(8) Pen˜a, D.; Pe´rez, D.; Guitia´n, E.; Castedo, L. Org. Lett. 1999, 1, 1555-
1557.
(9) For related Pd-catalyzed cycloadditions of arynes, see: (a) Pen˜a, D.;
Escudero, S.; Pe´rez, D.; Guitia´n, E.; Castedo, L. Angew. Chem., Int. Ed.
1998, 37, 2659-2661. (b) Pen˜a, D.; Pe´rez, D.; Guitia´n, E.; Castedo, L. J.
Am. Chem. Soc. 1999, 121, 5827-5828.
(10) Experimental procedure: To a solution of triflate 3 (106 mg, 0.27
mmol) and Pd2(dba)3‚CHCl3 (14 mg, 0.013 mmol) in CH3CN (5 mL) was
added finely powdered anhydrous CsF (122 mg, 0.8 mmol), and the mixture
was stirred at room temperature for 12 h. The resulting suspension was
filtered under vacuum. The solid obtained was washed with CH3CN (2 ×
1 mL) and Et2O (0.5 mL) and dissolved in CH2Cl2/H2O 9:1 (40 mL) The
organic layer was separated, dried over Na2SO4, filtered through a short
pad of SiO2 (CH2Cl2 as eluant), and concentrated under reduced pressure
to afford 1 (32 mg, 68%) as a yellow solid, only slightly soluble in common
organic solvents: 1H NMR (CDCl3, 500 MHz) δ 8.63 (d, J ) 8.0 Hz, 6H),
8.18 (d, J ) 8.2 Hz, 6H), 7.59 (dt, J ) 7.5, 1.2 Hz, 6H), 7.25 (dt, J ) 7.6,
1.0 Hz, 6H); MS, m/z (%) 528 (M+, 29), 264 (16), 261 (66), 254 (100);
HRMS for C42H24, calcd 528.1878, found 528.1869.
(11) The data that we initially gave for the spectrum in CDCl3, identical
to those previously reported,6,7 were obtained from a sample that had been
heated in solution for extended periods of time in an attempt to obtain a
satisfactory 13C NMR spectrum.
1630
Org. Lett., Vol. 2, No. 11, 2000