Cyclopentacorannulene: π-Facial stereoselective deuterogenation and determination of the bowl-to-bowl inversion barrier for a constrained buckybowl

Article abstract of DOI:10.1021/ja9521987

Attachment of an ethane or ethylene unit to the rim of corannulene produced relatively rigid bowls of dihydrocyclopenta- and cyclopentacorannulene, respectively. In contrast to the parent corannulenes, their inversion barriers are too high to be determined by the NMR coalescence methods. Due to the significant curvature of cyclopentacorannulene, deuterogenation is π-facial specific; both heterogeneous and homogeneous catalysis lead exclusively to exo-dideuteriocyclopentacorannulene 2a. Equilibration of the endo- and exo-isotopomers allowed the determination of ΔG? at 27.61-27.67 kcal/mol over the temperature range 52.1-99.3°C and the estimation of ΔH? (27.3 ± 0.7 kcal/mol) and ΔS? (-1.1 ± 0.2 eu) for the bowl-to-bowl inversion. The inversion barrier calculated at the HF/6-31G*//3-21G level (25.9 kcal/mol) compares well with the experimental result.

Full text of DOI:10.1021/ja9521987

J. Am. Chem. Soc. 1996, 118, 339-343
339
Cyclopentacorannulene: π-Facial Stereoselective
Deuterogenation and Determination of the Bowl-to-Bowl
Inversion Barrier for a Constrained Buckybowl
Andrzej Sygula, Atteye H. Abdourazak, and Peter W. Rabideau*
Contribution from the Department of Chemistry, Louisiana State UniVersity,
Baton Rouge, Louisiana 70803
ReceiVed July 5, 1995. ReVised Manuscript ReceiVed NoVember 6, 1995X
Abstract: Attachment of an ethane or ethylene unit to the rim of corannulene produced relatively rigid bowls of
dihydrocyclopenta- and cyclopentacorannulene, respectively. In contrast to the parent corannulenes, their inversion
barriers are too high to be determined by the NMR coalescence methods. Due to the significant curvature of
cyclopentacorannulene, deuterogenation is π-facial specific; both heterogeneous and homogeneous catalysis lead
exclusively to exo-dideuteriocyclopentacorannulene 2a. Equilibration of the endo- and exo-isotopomers allowed
the determination of ∆G‡ at 27.61-27.67 kcal/mol over the temperature range 52.1-99.3 °C and the estimation of
∆H‡ (27.3 ( 0.7 kcal/mol) and ∆S‡ (-1.1 ( 0.2 eu) for the bowl-to-bowl inversion. The inversion barrier calculated
at the HF/6-31G*//3-21G level (25.9 kcal/mol) compares well with the experimental result.
The discovery of buckminsterfullerene (C60) and related
carbon cages1 has led to a general interest in the properties and
behavior of curved carbon systems. An examination of the
buckminsterfullerene surface leads to the recognition of many
familiar carbon frameworks that, if removed from C60 and the
dangling bonds occupied with hydrogen, would produce com-
mon hydrocarbons. The lower molecular weight hydrocarbons
constructed in this way are planar, however, and it is not until
C20 is reached that curvature remains as a feature of the structure.
Beyond C20, the bowl-shaped geometries are expected to be
quite rigid, and we have referred to these fullerene-related,
curved surface aromatic hydrocarbons as “buckybowls”.2
The carbon framework of corannulene3-6 (1) represents the
polar cap of buckminsterfullerene. Corannulene is a fascinating
molecule, initially prepared by Barth and Lawton,3 that has now
become more accessible due to some elegant synthetic work.4,5
In spite of its substantial curvature,6 corannulene is surprisingly
flexible, undergoing rapid bowl-to-bowl inversion in solution7
as evidenced by the dynamic NMR behavior of several
derivatives. Hence, due to the bowl-shaped geometry, the
attachment of suitable substituents such as dimethylcarbinol,
isopropyl, and benzyl (1-R), leads to the observation of two
sets of diastereotopic proton signals at low temperatures (-31
to -64 °C). Warming produces coalescence, and inversion
barriers in the range of ∆G‡ ) 10-11 kcal/mol have been
determined.
X Abstract published in AdVance ACS Abstracts, December 15, 1995.
(1) (a) Kroto, H. W.; Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley,
R. E. Nature 1985, 318, 162. (b) Kra¨tschmer, W.; Lamb, L. D.; Fostirop-
oulos, K.; Huffman, D. R. Nature 1990, 347, 354. (c) Kroto, H. W.; Allaf,
A. W.; Balm, S. P. Chem. ReV. 1991, 91, 1213-1235. (d) Hammond, G.
S., Kuck, V. J., Eds. Fullerenes; ACS Symposium Series 481; American
Chemical Society: Washington, DC, 1992. (e) Acc. Chem. Res. 1992, 25
(a special issue on buckminsterfullerenes). (f) Taylor, R.; Walton, D. R.
M. Nature 1993, 363, 685. (g) Hirsch, A. Angew. Chem., Int. Ed. Engl.
1993, 32, 1138. (h) Billups, W. E., Ciufolini, M. A., Eds. Buckminster-
fullerenes; VCH Publishers, New York, 1993. (i) Kroto, H. W., Fischer, J.
E., Cox, D. E., Eds. The Fullerenes; Pergamon: Oxford, 1993. (j) Koruga,
D.; Hamerloff, S.; Withers, J.; Loutfy, R.; Sundareshan, M. Fullerene C60.
History, Physics, Nanobiology, Nanotechnology; Elsevier: Amsterdam,
1993. (k) Kroto, H. W., Walton, D. R. M., Eds. The Fullerenes. New
Horizons for the Chemistry, Physics and Astrophysics of Carbon; Cambridge
Univ. Press: Cambridge, 1993.
Hence the relatively low barrier to bowl-to-bowl inversion
in corannulene raises the question as to how large a section
from the buckminsterfullerene surface is necessary to produce
a hydrocarbon with a constrained or rigid-bowl structure. The
work described herein demonstrates that rigid structures will
be encountered quickly with only minor expansion of the
corannulene carbon framework.
Results and Discussion
In an effort to explore the flexibility of the bowl-shaped
geometry beyond corannulene, we undertook the synthesis of
dihydrocyclopentacorannulene (2).9 We followed the same
(2) Rabideau, P. W.; Abdourazak, A. H.; Folsom, H. E.; Marcinow, Z.;
Sygula, A.; Sygula, R. J. Am. Chem. Soc. 1994, 116, 7891.
(3) Barth, W. E.; Lawton, R. G. J. Am. Chem. Soc. 1966, 88, 381. Barth,
W. E.; Lawton, R. G. J. Am. Chem. Soc. 1971, 93, 1730.
(4) Scott, L. T.; Hashemi, M. M.; Meyer, D. T.; Warren, H. B. J. Am.
Chem. Soc. 1991, 113, 7082.
(5) See also: Borchardt, A.; Fuchicello, A.; Kilway, K. V.; Baldridge,
K. K.; Siegel, J. S. J. Am. Chem. Soc. 1992, 114, 1921.
(7) Scott, L. T.; Hashemi, M. M.; Bratcher, M. S. J. Am. Chem. Soc.
1992, 114, 1920.
(8) Sygula, A.; Rabideau, P. W., unpublished results.
(9) For a preliminary account, see: Abdourazak, A. H.; Sygula, A.;
Rabideau, P. W. J. Am. Chem. Soc. 1993, 115, 3010.
(6) Hanson, J. C.; Nordman, C. E. Acta Crystallogr. B 1976, B32, 1147.
0002-7863/96/1518-0339$12.00/0 © 1996 American Chemical Society

340 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Sygula et al.
procedure as the Scott group,4 but with an additional five-
membered ring incorporated into the starting material (Scheme
1). This procedure was successful except that hydrogen was
Scheme 1
Figure 1. POAV1 pyramidalization angles for carbon atoms in 3 based
on the ab initio calculated geometry (right) and crystal structure
geometry (left). CH carbon angles for the latter are omitted since
hydrogen atoms were placed in calculated positions in the crystal
structure analysis.10a
Figure 2. Crystal packing in 3 as viewed approximately along the
crystallographic a-axis.
lost during the final pyrolysis step, producing unsaturated 3
rather than 2, and some corannulene was also produced as a
result of loss of the ethane bridge. The separation of 3 from
corannulene (1) proved to be quite difficult, and so in an attempt
to minimize the amount of 1 formed, the external five-membered
ring was unsaturated in the hope that an ethene unit would be
more stable to the pyrolysis conditions (Scheme 2). In fact,
curvature as demonstrated by x-ray crystal structure de-
termination10a and predicted by ab initio calculations.10b Figure
1 shows the POAV111 pyramidalization angles for the carbon
atoms in 3 calculated from both the crystal structure and ab
initio HF/6-31G* geometries. These angles are consistently
larger than those for 1sindicating the carbon atoms to be more
pyramidalizedssince the POAV1 angles for corannulene are
8.4, 3.7, and 1.5° for the core, quat rim, and CH rim carbon
atoms, respectively. Also, the pyramidalization of the carbon
atom in 3 in the region of maximum curvature is quite
comparable with the curvature of buckminsterfullerene (11.6°).11
The very good agreement of the POAV1 angles in 3 calculated
from both the experimental and theoretical geometries supports
the validity of the theoretical model used.
Scheme 2
Crystal structure determination for 3 also revealed an interest-
ing long-range packing pattern in the solid state.10a The bowl-
shaped molecules are stacked in a concave-convex fashion
(Figure 2) in which there are 24 intermolecular C‚‚‚C distances
shorter than 3.8 Å for every two molecules in the crystal. While
such stacking is absent in the crystal of corannulene,6 and only
a few crystal structure determinations of curved surface
this afforded even poorer results since corannulene was still
produced, and there was also an increase in ring open products
as well as some polymerization of the unsaturated precursor in
the pyrolysis apparatus before it reached the hot zone. Thus,
the pathway described in Scheme 1 remains the method of
choice for the synthesis of 3 at the moment. The hydrogenation
of 3 to produce 2 proceeded quantatively at 1 atm in the presence
of P-2 nickel.
(10) (a) Sygula, A.; Folsom, H. E.; Sygula, R.; Abdourazak, A. H.;
Marcinow, Z.; Fronczek, F. R.; Rabideau, P. W. J. Chem. Soc., Chem.
Commun. 1994, 2571. (b) Sygula, A.; Rabideau, P. W. J. Chem. Soc., Chem.
Commun. 1994, 1497.
(11) (a) Haddon, R. C.; Scott, L. T. Pure Appl. Chem. 1986, 58, 137.
(b) Haddon, R. C. Acc. Chem. Res. 1988, 21, 243. (c) Haddon, R. C. J.
Am. Chem. Soc. 1990, 112, 3385. (d) Haddon, R. C. Science 1993, 261,
1543.
The incorporation of an additional five-membered ring onto
the rim of corannulene produces a significant increase in

Cyclopentacorannulene
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 341
Table 1. Rate Constants (k), Half-Life Times (τ1/2), and Free
Enthalpies of Activation (∆G‡) Determined for the Equilibration
2a-2b
t (°C)
106k (s-1
)
τ
1/2 (min)
∆G‡ (kcal/mol)
99.3
94.0
78.9
52.1
453 ( 6
12.8
21.6
114.6
3024.2
27.67 ( 0.02
27.65 ( 0.02
27.65 ( 0.02
27.61 ( 0.02
267 ( 5
50.4 ( 0.8
1.91 ( 0.02
expected to be shifted upfield relative to the convex hydrogens.
This assignment is also supported by NOE results.13 Irradiation
of the aromatic hydrogens proximal to the benzylic hydrogens
(Har at 7.0 ppm) resulted in ca. twice the enhancement of the
high-field aliphatic signal, and so the signal at 2.41 ppm
remaining after deuterogenation can be assigned as concave or
endo since, as calculated at the HF/3-21G level of theory, the
Har-Hendo distance is 2.92 Å, as compared to 3.31 Å for Har-
Hexo. Thus, heterogeneous deuterogenation is π-facial specific,
occurring exclusively on the convex side of 3. Moreover, exo
π-facial specificity was also observed under homogeneous
conditions using the Willkinson type [Rh(norbornadiene)-
(PPh3)2]+BF4- catalyst.14 Again, only endo benzylic hydrogens
at 2.41 ppm could be detected for deuterogenation of 3 with
this catalyst.
Clearly the significant curvature of 3, coupled with the
absence of rapid bowl-to-bowl inversion, facilitates exofacial
deuterogenation. Apparently convex-face complexation of 3
with the metal surface in the heterogeneous catalysis is strongly
preferred over the concave arrangement, probably due to better
overlap of the π-orbitals of the external five-membered ring
with the orbitals of the metal in the complex. Similarly, convex
complexation of the homogeneous catalyst with the external
five-membered ring in 3 is favored over concave. This was an
interesting case since it allowed for the possibility of the catalyst
to be complexed within the bowl. Evidently, there is not enough
space inside the bowl to accommodate this bulky rhodium
complex.
Figure 3. Equilibration of dideuteriocyclopentacorannulene isoto-
pomers 2a and 2b at 78.9 °C in benzene-d6 as monitored by 1H NMR
of the alicyclic region.
hydrocarbons have been done, it is not clear whether concave-
convex stacking in 3 is an exception rather than common
behavior for such systems. This type of stacking, however, is
expected to maximize attractive van der Waals interactions.
Because of its symmetry, bowl-to-bowl inversion in 3 could
not be observed by NMR without appropriate substitution. In
contrast, 2 has two sets of aliphatic hydrogens that would
exchange with bowl-to-bowl inversion. However, the two sets
of signals representing the distinguishable exo (convex) and
endo (concave) hydrogens in 2 were observed up to 135 °C
without change, and a lower limit for bowl-to-bowl inversion
was estimated at ca. 26 kcal/mol at 127 °C by the spin
polarization transfer method.9 Thus the inversion in 2 is too
slow for determination of the barrier by NMR coalescence
methods.
While a suitably placed substituent would render 2 chiral,
and the bowl-to-bowl barriersa racemization processscould
be determined by the thermal decay of optical activity for one
of the resolved enantiomers, the relatively small amounts of 2
available made this process rather unattractive. Hence we hoped
that the considerable curvature of 3 might lead to π-facial
stereoselectivity in hydrogenation since, if so, deuterogenation
of 3 would provide an uneven distribution of endo and exo
deuterogenated products. In principle, this would allow deter-
mination of the inversion barrier by following the equilibration
While exo-dideuteriocyclopentacorannulene (2a) showed no
change on standing at room temperature for several days, heating
the sample caused the signal at 2.41 ppm to diminish with
concurrent growth of a signal at 3.26 ppm (Figure 3), indicating
that bowl-to-bowl inversion does indeed take place in this
system. We determined rate constants for the equilibration in
1
of the two isotopomers by H NMR.
Heterogeneous deuterogenation of 3 with P-2 Ni as a catalyst
led to the formation of a single dideuterio product as shown by
the presence of a broadened singlet at 2.41 ppm (in benzene-
d6) from two benzylic protons together with a little of the
partially hydrogenated product (Figure 3, bottom trace).12 No
signal was detected in the vicinity of 3.3 ppm that would indicate
the presence of a second isomer. We assigned the signal at
2.41 ppm to endo (concave) protons since they have greater
proximity to the center of the nearby benzene rings and are
the temperature range from 52.1 to 99.3 °C, and the results are
presented in Table 1.
The introduction of a single, additional five-membered ring
to the rim of corannulene increases the barrier for inversion
(12) The partially hydrogenated byproduct (ca. 5% by 1H NMR) is most
probably formed by the addition of HD, a contaminant in the D2 gas used.
1H NMR showed low-intensity multiplets centered at 2.97 and 3.78 ppm
(in CDCl3), and a DEPT-135 experiment (13C NMR) showed two minor
CH2 carbon atoms at ca. 31.8 ppm in addition to the strong CHD triplet at
31.43 ppm.
(13) An NOE experiment was conducted on the equilibriated mixture
of exo and endo dideuterio isomers.
(14) Deuterogenation was achieved according to the procedure developed
by Osborn.15 The cationic rhodium catalyst was kindly provided by Barry
Misquitta and Professor George Stanley, Louisiana State University.

342 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Sygula et al.
more than 2-fold! It is also evident from Table 1 that ∆G‡
shows little temperature dependency since ∆∆G‡ throughout
the entire temperature range is 0.06 kcal/mol, only 3 times higher
than the standard deviation. Apparently ∆S‡ is small; values
of ∆S‡ ) -1.1 ( 0.2 eu16 and ∆H‡ ) 27.3 ( 0.7 kcal/mol are
estimated from the data of Table 1.
At the ab initio level of theory, bowl-shaped 2 is a minimum-
energy conformation with all positive vibrational frequencies
and EHF ) -840.074 07 au (HF/6-31G*//3-21G). The planar
conformer represents the transition state (TS) for the bowl
inversion process since it exhibits one imaginary frequency
(140i), and the eigenvector associated with this frequency
distorts the planar structure toward the bowl-shaped minimum-
energy conformation. At the HF/6-31G*//3-21G level, the
energy of the planar TS is 25.9 kcal/mol higher than the energy
of the bowl-shaped conformer. Thus, the computational results
compare favorably with the experimental determination of the
inversion barrier in 2. Furthermore, a previously reported
calculation for 3 at the same level of theory yielded a value of
28.8 kcal/mol,10b suggesting that the barrier to bowl-to-bowl
inversion for 3 is even higher than 2.
In conclusion, introduction of the additional five-membered
ring to the corannulene carbon framework increased significantly
both the curvature and rigidity of the system. Deuterogenation
of 3 was found to be π-facial specific due to the high degree of
curvature present in the hydrocarbon surface; both heterogeneous
and homogeneous catalysis provided the exo-dideuterated
product 2a. Equilibration of benzene solutions of 2a allowed
the determination of ∆G‡ for the bowl-to-bowl inversion
(27.61-27.67 kcal/mol over the temperature range 52.1-99.3
°C) and an estimation of ∆H‡ (27.3 ( 0.7 kcal/mol) and ∆S‡
(-1.1 ( 0.2 eu).
mixture was stirred overnight, and filtered and the dark-silver solid
washed with methanol. After drying, 3.56 g (40% yield) of crude 5
was obtained which was subsequently used without further purification.
Crystallization from DMF gave a dark solid with mp 235-240 °C dec.
1H NMR (CDCl3, 200.13 MHz): δ 8.6 (d, J ) 7.4 Hz, 2H), 7.49 (d,
J ) 7.4 Hz, 2H), 3.58 (s, 4H), 2.64 (s, 6H). HRMS: calcd 314.0942,
found 314.0933.
Preparation of 1,1′-(1,2-dihydrocyclopenta[cd]fluoranthene-5,8-
diyl)bisethanone (6). A solution of 8.6 g (28.34 mmol) of crude 5
and 5.2 mL (48.18 mmol) of bicyclo[2.2.1]hepta-2,5-diene in 150 mL
of butanol was refluxed overnight. After cooling, the brown solid was
separated and chromatographed on silica gel with hexane/ethyl acetate
(10:1 then 5:1) or DCM to provide 5.6 g (64%) of yellowish solid.
Recrystallization from ethanol-toluene (ca. 4:1) gave yellow needles,
mp 249-251 °C dec. 1H NMR (CDCl3, 250.13 MHz): δ 8.49 (d, J )
7.5 Hz, 2H), 7.7 (s, 2H), 7.48 (d, J ) 7.5 Hz, 2H), 3.53 (s, 4H), 2.8 (s,
6H). 13C NMR (CDCl3, 62.89 MHz): δ 29.69, 32.27, 120.91, 126.03,
128.53, 129.43, 132.04, 136.31, 137.44, 138.86, 148.17, 201.45. MS:
m/z 312 (100, M+), 297 (100), 269 (24), 254 (19), 226 (42). Anal.
Calcd for C22H16O2: C, 84.59; H, 5.16; O, 10.25. Found: C, 84.31;
H, 5.16. HRMS: calcd 312.1150, found 312.1152.
Preparation of 5,8-Bis(1-chloroethenyl)-1,2-dihydrocyclopenta-
[cd]fluoranthene (7). A mixture of 1.08 g (3.46 mmol) of 7 and 2.52
g (12.11 mmol) of PCl5 was refluxed for 3 h in 40 mL of benzene.
After cooling, the mixture was poured onto ice, the organic layer was
washed with water and sodium carbonate solution and dried over
magnesium sulfate and the solvent was evaporated. After purification
by column chromatography on silica gel with cyclohexane as eluant,
0.6 g of yellow solid was obtained (50% yield). Recrystallization from
methanol provided yellow needles. Mp 144-146 °C. 1H NMR
(CDCl3, 250.13 MHz): δ 8.3 (d, 2H, J ) 7 Hz), 7.46 (d, 2H, J ) 7
Hz), 7.37 (s, 2H), 5.92 (d, 2H, J ) 1.2 Hz), 5.8 (d, 2H, J ) 1.2 Hz),
3.51 (s, 4H). 13C NMR (CDCl3, 62.89 MHz): δ 32.27, 117.02, 120.88,
126.18, 127.07, 130.14, 131.32, 135.34, 136.39, 137.31, 138.30, 147.01.
MS: m/z 350 (44), 348 (69, M+), 313 (42), 276 (100), 263 (15), 250
(57), 138 (69). Anal. Calcd for C22H14Cl2: C, 75.66; H, 4.04; Cl,
20.30. Found: C, 75.21; H, 4.06; Cl, 20.11. HRMS: calcd 348.0473,
found 348.0458.
Experimental Section
Chemicals were purchased from Aldrich Chemicals. Acenaphthene
was recrystallized from ethanol, and N-bromosuccinimide from water.
2,4,6-Heptanetrione17 and 1,2-diketopyracene (4)18 were prepared by
literature methods. Melting points are reported uncorrected. HRMS
were performed at the LSU Mass Spec Facility. Elemental analyses
were performed by Oneida Research Services, Inc., Whitesboro, N.Y.
Ab initio calculations were performed using the Gaussan92/DFT
package.19a Geometry optimization and vibrational frequency calcula-
tions were done at the HF/3-21G level.19b Single-point calculations
were performed with 3-21G geometries and the 6-31G* basis set19c
(HF/6-31G*//3-21G level).
Preparation of 5,7-Diacetyl-1,2-dihydro-6H-dicyclopent[a,fg
]acenaphthylen-6-one (5). A solution of 5.8 mL (41.7 mmol) of
triethylamine in 20 mL of methanol was added dropwise to a stirred
solution of 5.9 g (28.36 mmol) of 1,2-diketopyracene (4) and 6.8 g
(47.53 mmol) 2,4,6-heptanetrione in 50 mL of methanol. The reaction
Pyrolysis of 7. A 985 mg sample of 7 was pyrolyzed in batches of
80 mg each at 1000 °C under a slow bleed of nitrogen at 1.5 mmHg.
After each run (3 h), the pyrolysis product was washed out of the trap
and the elbow with DCM and combined. The solvent was evaporated
and the crude material flash chromatographed on silica gel with hexane
as eluant to give a mixture of 3 and 1 (7:3, by GC/MS). Yields varied
from 10 to 15%. The mixture was chromatographed again (silica gel,
cyclohexane) and then recrystallized three times from ethanol and once
from ether to provide pure 3 as orange platelets. 1H NMR (CDCl3,
200.13 MHz): δ 6.49 (s, 2H), 7.31 (s, 2H), 7.38 (s, 2H), 7.44 (d, 2H,
J ) 9 Hz), 7.50 (d, 2H, J ) 9 Hz). 13C NMR (CDCl3, 100.61 MHz):
δ 124.17, 126.46, 127.26, 128.28, 128.42, 129.73, 137.51, 137.89,
138.16. MS: m/z (relative intensity) 274 (100, M+), 272 (18), 137
(20), 136 (18). Final structure proof was accomplished by X-ray
crystallography.10a
Preparation of 8,9-dihydrodibenzo[ghi,mno]cyclopenta[cd]-
fluoranthene (2). This procedure follows the methodology of Brown
and Ahuja.20 Nickel acetate tetrahydrate (0.07 g, 0.29 mmol) was
dissolved in 3 mL of 95% ethanol and placed in a 50 mL three-neck,
round-bottomed flask equipped with a magnetic stirrer. The system
was purged and filled with hydrogen three times. Then 0.3 mL of a
solution prepared by mixing 0.25 g of sodium borohydride, 6 mL of
95% ethanol, and 0.31 mL of 2 N aqueous sodium hydroxide was added,
followed by the addition of 3 mg of 3 dissolved in 1 mL of DCM.
After 15 min, the solution was filtered through a Celite pad and
evaporated. The yield was quantitative by GC, and the residue was
chromatographed on silica gel with hexane as eluant to afford a yellow
solid. 1H NMR (CDCl3, 200.13 MHz): δ 2.98, 3.39 (4H), 7.34 (s,
2H), 7.70 (s, 4H), 7.72 (s, 2H). 13C NMR (CDCl3, 100.61 MHz): δ
31.83, 121.74, 126.32, 126.96, 127.68, 130.31, 137.19, 137.53, 138.71,
139.41, 145.04, 147.51. MS: m/z 276 (100, M+), 275 (21), 274 (34),
138 (27), 137 (32), 136 (18). HRMS: calcd 276.0939, found 276.0937.
(15) Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1976, 98, 2134.
(16) ∆S‡ for the inversion of dihydrocyclopentacorannulene 2 is expected
to be ca. -2.5 eu for symmetry reasons. The bowl-shaped 2 is of Cs
symmetry with symmetry number 1, while the planar TS is of C2V symmetry
(symmetry number 2). Thus, there is an additional -R ln 2 (or -1.38 eu)
symmetry contribution to ∆S‡. In the case of 2a and 2b, both the minimum-
energy conformers and the planar TS have the same Cs symmetry.
(17) Feist, F.; Belart, H. Chem. Ber. 1895, 28, 1823.
(18) Trost, B. A. J. Am. Chem. Soc. 1969, 91, 918.
(19) (a) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Wong, M. W.; Foresman, J. B.; Robb, M. A.; Head-Gordon,
M.; Reploge, E. S.; Gomperts, R.; Andres, J. L.; Raghavachari, K.; Binkley,
J. S.; Gonzalez, C.; Martin, R. L.; Fox, D. J.; Defrees, D. J.; Baker, J.;
Stewart, J. J. P.; Pople, J. A. GAUSSIAN92/DFT, ReVision F.4; Gaussian,
Inc.: Pittsburgh, PA, 1993. (b) 3-21G basis set: Binkley, J. S.; Pople, J.
A.; Hehre, W. J. J. Am. Chem. Soc. 1980, 102, 939. Gordon, M. J.; Binkley,
J. S.; Pople, J. A.; Pietro, W. J.; Hehre, W. J. J. Am. Chem. Soc. 1982, 104,
2797. (c) 6-31G* basis set: Hariharan, P. C.; Pople, J. A. Chem. Phys.
Lett. 1972, 66, 217. Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J.
S.; Gordon, M. S.; DeFries, D. J.; Pople, J. A. J. Chem. Phys. 1982, 77,
3654.
(20) Brown, C. A.; Ahuja, Y. K. J. Org. Chem. 1973, 38, 2276.

Cyclopentacorannulene
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 343
Preparation of 5,8-Bis(1-chloroethenyl)cyclopenta[cd]fluoranthene
(8). A mixture of 428 mg (1.23 mmol) of 7 and 490 mg (2.75 mmol)
of NBS in 20 mL of benzene was refluxed for 6 h. The solvent was
evaporated, the residue was taken in DCM and washed with water,
and the organic layer was dried and evaporated. Chromatography on
silica gel with DCM/hexane (10:1) gave 300 mg (57%) of the dibromide
of 7. 1H NMR (CDCl3, 250.13 MHz): δ 8.34 (d, 2H, J ) 7.5 Hz),
7.66 (d, 2H, J ) 7.5 Hz), 7.36 (s, 2H), 6.03 (bs, 2H), 5.89 (bs, 2H),
5.75 (bs, 2H). 13C NMR (CDCl3, 62.89 MHz): δ 55.90, 117.77, 124.16,
127.21, 128.57, 132.21, 133.86, 136.32, 137.75, 141.03.
The above dibromide (224 mg, 0.53 mmol) was refluxed overnight
with 1.1 g (7 mmol) of potassium iodide in 30 mL of acetone. After
cooling, the mixture was filtered and the solid residue washed with
acetone. The filtrate was evaporated and the red solid was chromato-
graphed on silica gel with hexane affording 89 mg (49%) of a reddish
solid that decomposes over 220 °C. 1H NMR (CDCl3, 250.13 MHz):
δ 5.65 (d, 2H, J ) 1.33 Hz), 5.76 (d, 2H, J ) 1.33 Hz), 6.52 (s, 2H),
7.11 (d, 2H, J ) 7.25 Hz), 7.51 (d, 2H, J )7.25 Hz). 13C NMR (CDCl3,
62.89 MHz): δ 29.92, 117.29, 125.42, 125.88, 129.08, 132.21, 136.37,
137.58, 141.53. MS: m/z 348 (29), 346 (44, M+), 311 (23), 276 (56),
275 (51), 274 (60), 250 (16), 155 (18), 137 (100), 124 (32). HRMS:
calcd 346.0316, found 346.0305.
Heterogeneous Deuterogenation of Dibenzo[ghi,mno]cyclopenta-
[cd]fluoranthene (3). This was achieved by the method described
above except that the hydrogen gas was replaced with deuterium (99.6%
purity) and sodium borohydride was replaced with sodium borodeu-
terate: yield quantitative. 1H NMR (250 MHz, CDCl3): δ 2.95 (s,
2H), 7.34 (s, 2H), 7.70 (s, 4H), 7.72 (s, 2H). 1H NMR (400.13 MHz,
C6D6): δ 2.41 (s, 2 H), 7.00 (s, 2 H), 7.48 (s, 4H), 7.51 (s, 2H). 13C
NMR (100.61 MHz, CDCl3) δ 31.43, 121.80, 126.32, 126.94, 127.68,
130.28, 137.18, 137.53, 138.69, 139.39, 145.05, 147.51. MS: m/z 278
(100), 277 (67), 276 (37), 275 (24), 139 (45), 138 (54), 137 (53), 136
(18), 125 (20).
with deuterium gas three times. The orange color of the solution faded
in a few seconds. The solution was stirred vigorously at room
temperature and at constant pressure (1 atm) for 10 min. Next, 20 mg
(0.073 mmol) of 3 in 2 mL of THF was injected into the solution, and
the reaction mixture was stirred for 1 h. The solvent was removed
and the black residue chromatographed on silica gel with hexane to
give 15 mg of the dideuterated product which showed a 1H NMR
spectrum identical to that of 2a obtained by the heterogeneous catalysis
experiment.
Kinetics of the Equilibration of the Isotopomers 2a and 2b. The
NMR tubes were charged with ca. 4 mg portions of 2a dissolved in
benzene-d6, then degassed, and sealed under vacuum. The samples
were kept in a thermostat for the appropriate period of time at the
desired temperature ((0.1 °C), then removed, and quickly cooled in
ice-water. The progress of equilibration was monitored by 1H NMR
at ambient temperature with a relaxation delay of 20 s. The benzylic
region of each spectrum was carefully phase- and baseline-corrected
before integration. The rate constants (k’s) for the reversible equilibra-
tion were determined by regression analysis using the equation 2kt )
ln[a/(a - 2x)] where a is the initial concentration of 2a and x is the
concentration of 2b at time t. Correlation coefficients of the linear
regressions were 0.998 or higher. The standard deviations (SD’s) of
the slopes of the resulting regression lines were used as the estimates
for the SD’s of the rate constants. ∆G‡’s were calculated from the
Eyring equation, and SD’s were estimated by the error propagation
method.
Acknowledgment. This work was supported by the Division
of Chemical Sciences, Office of Basic Energy Sciences, of the
U.S. Department of Energy, and by SNCC (Louisiana State
University) for allocation of computer time. We also thank
Renata Sygula and Haskell E. Folsom for help with the synthesis
of 3.
Homogeneous deuterogenation of Dibenzo[ghi,mno]cyclopenta-
[cd]fluoranthene (3). [Rh(NBD)(PPh3)2]+BF4- (60 mg, 0.074 mmol)
in 10 mL of dry THF in a 25 mL round-bottomed flask was flushed
JA9521987