High-yield double cyclopropanation of benzene

Full text of DOI:10.1021/jo010277x

J . Org. Chem. 2001, 66, 4945-4949
4945
While investigating a completely different facet of Rh-
stabilized carbenoid chemistry, namely carbenoid N-H
insertion, we had occasion to try the Rh2(OAc)4-catalyzed
reaction of dimethyl diazomalonate, (MeOOC)2CdN2
(“DDM”), with hexamethyldisilazane (HMDS) in reflux-
ing benzene. We expected this reaction to lead to 6 via
insertion of the DDM-derived carbenoid into the N-H
bond of HMDS (eq 2). Surprisingly, silica gel column
High -Yield Dou ble Cyclop r op a n a tion of
Ben zen e
M. Yang, T. R. Webb, and P. Livant*
Department of Chemistry, Auburn University,
Auburn, Alabama 36849-5312
livanpd@auburn.edu
Received March 12, 2001
In tr od u ction
The thermal or photochemical reaction of ethyl diaz-
oacetate with benzene to give (via norcaradiene 1) a
mixture of isomeric cycloheptatrienes (2-5) (eq 1) is the
prototypical example of the Bu¨chner reaction, a reaction
that has been known for over 100 years.1 The daunting
chromatographic workup afforded two fractions, A and
B, neither of which contained trimethylsilyl groups. The
reaction was repeated in the absence of HMDS and the
identical products were obtained. Therefore, HMDS was
superfluous; all products resulted from a Bu¨chner reac-
tion of DDM with the solvent, benzene. The details of
this reaction are reported herein.
Resu lts a n d Discu ssion
Fraction A was identified as a mixture of 7, 8, and 9
(eq 3), based on comparison with reported 1H NMR
spectra.5-7 Compounds 7 and 8 have been shown to be
complexity of the product mixtures was reduced or
eliminated with the advent of metal catalysts, at first
copper-based, then in the early 1980s Rh2(OAc)4, which,
along with its analogues, has become the catalyst of
choice for this reaction.2 Subsequently, Rh2(OAc)4-
catalyzed cyclopropanations of aromatics, especially in-
tramolecular cyclopropanations, have enjoyed a certain
popularity due to the regio- and stereoselectivity which
can now be achieved.3 The related sequential cyclopro-
panation-Cope rearrangement methodology that has
been applied fruitfully to a variety of synthetic challenges
includes examples of Bu¨chner-type cyclopropanations of
benzenes, pyrroles, and furans.4
* To whom correspondence should be addressed. Fax: 334 844 6959.
(1) (a) Doyle, M. P.; McKervey, M. A.; Ye, T. Modern Catalytic
Methods for Organic Synthesis with Diazo Compounds; Wiley: New
York, 1998; ISBN 0-471-13556-9, Chapter 6. (b) Maas, G. Top. Curr.
Chem. 1987, 137, 75-253 (c) Davies, H. M. L. In Comprehensive
Organic Synthesis; Trost, B. M., Ed.; Pergamon: Oxford, 1991; Vol. 4,
Chapter 4.8.
(2) (a) Anciaux, A. J .; Demonceau, A.; Hubert, A. J .; Noels, A. F.;
Petiniot, N.; Teyssie´, P. J . Chem. Soc., Chem. Commun. 1980, 765-
766. (b) Anciaux, A. J .; Demonceau, A. F.; Noels, A. J .; Hubert, A. J .;
Warin, R.; Teyssie´, P. J . Org. Chem. 1981, 46, 873-876.
(3) (a) Kennedy, M.; McKervey, M. A.; Maguire, A. R.; Tuladhar, S.
M.; Twohig, M. F. J . Chem. Soc., Perkin Trans. 1 1990, 1047-1054.
(b) Duddeck, H.; Ferguson, G.; Kaitner, B.; Kennedy, M.; McKervey,
M. A.; Maguire, A. R. J . Chem. Soc., Perkin Trans. 1 1990, 1055-
1063. (c) Cordi, A. A.; Lacoste, J .-M.; Hennig, P. J . Chem. Soc., Perkin
Trans. 1 1993, 3-4. (d) Manitto, P.; Monti, D.; Speranza, G. J . Org.
Chem. 1995, 60, 484-485. (e) Maguire, A. R.; Buckley, N. R.; O’Leary,
P.; Ferguson, G. J . Chem. Soc., Chem. Commun. 1996, 2595-2596. (f)
Manitto, P.; Monti, D.; Zanzola, S.; Speranza, G. J . Org. Chem. 1997,
62, 6658-6665. (g) Moody, C. J .; Miah, S.; Slawin, A. M. Z.; Mansfield,
D. J .; Richards, I. C. J . Chem. Soc., Perkin Trans. 1 1998, 4067-4075.
(h) Maguire, A. R.; Buckley, N. R.; O’Leary, P.; Ferguson, G. J . Chem.
Soc., Perkin Trans. 1 1998, 4077-4091. (i) Doyle, M. P.; Ene, D. G.;
Forbes, D. C.; Pillow, T. H. J . Chem. Soc., Chem. Commun. 1999,
1691-1692.
in rapid equilibrium, with Keq ) 0.30 at the temperature
of refluxing benzene and Keq ) 0.31 at room tempera-
ture.6 The ratio of [7 + 8] to 9 depended somewhat on
the duration of reflux and was in the range [7 + 8]/9 )
2.3-3.0 for reflux times of 4-8 h. These products are
unremarkable, inasmuch as they have been reported
before for thermal and photochemical cases of the Bu¨ch-
ner reaction of DDM with benzene.8
By contrast, fraction B was indeed remarkable. It was
1
identified as bis-cyclopropanation product 10 by H and
13C NMR, mass spectrometry, elemental analysis, IR, and
(4) Davies, H. M. L.; Smith, H. D.; Hu, B.; Klenzak, S. M.; Hegner,
F. J . J . Org. Chem. 1992, 57, 6900-6903.
(5) Berson, J . A.; Hartter, D. R.; Klinger, H.; Grubb, P. W. J . Org.
Chem. 1968, 33, 1669-1671.
(6) Go¨rlitz, M.; Gu¨nther, H. Tetrahedron 1969, 25, 4467-4480.
(7) Ledon, H.; Linstrumelle, G.; J ulia, S. Bull. Soc. Chim. Fr. 1973,
2065-2071.
10.1021/jo010277x CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/20/2001

4946 J . Org. Chem., Vol. 66, No. 14, 2001
Notes
ultimately, X-ray crystallography. Compound 10 is the
major product of this reaction, being formed in 58% yield.
The use of Rh2(O2CCF3)4 instead of Rh2(OAc)4 as the
catalyst led to a vastly different product distribution: [7
+ 8], 64%; 9, 32%; 10, 4%. The low yield of double-
cyclopropanation product 10 obtained with Rh2(O2CCF3)4
is comparable to other carbenoid reactions with aromat-
ics, in which double-cyclopropanation is almost never
found. Despite the long history of the Bu¨chner reaction,
double cyclopropanation has been reported in only the
following few instances of which we are aware: (a) the
copper-bronze-catalyzed reaction of methyl diazoacetate
with benzene, which gave 11 in 3% yield;9 (b) the solvent-
free thermal reaction of naphthalene with excess DDM,
which gave 12 in 5-15% yield;10 (c) the thermal addition
of ethyl diazoacetate to naphthalene and two symmetrical
dimethylnaphthalenes, which gave 13a and 13b in 14-
17% yield (13a /13b ) 2.6),11 13c in 11% yield,12 and 13d
in 19% yield;13 (d) the CuCl-catalyzed reaction of diaz-
omethane with N-carbomethoxypyrrole, which gave 14a
in 12% yield and a trace of 14b, and the CuBr-catalyzed
reaction of ethyl diazoacetate with N-carbomethoxypyr-
role, which gave 14c in 5% yield;14 and (e) the Rh2(OAc)4-
catalyzed reaction of diethyl (Z)-2-methoxy-4-diazo-2-
pentenedioate with N-carbomethoxypyrrole, which gave
14d in 33% yield.15 The formation of diadduct 15 from
CuSO4-catalyzed addition of methyl diazoacetate to an-
thracene has been mentioned, but a yield was not given.16
Cases (d) and (e) are included here, even though N-
carbomethoxypyrrole is known to be anomalous in its
reactivity toward carbenoids, resembling more a diene
than a typical aromatic.14 Therefore, carbenoid double
cyclopropanation of an aromatic substrate is a rarity, and
a 58% yield of such a product is unprecedented.
with DDM and Rh2(OAc)4 in refluxing CCl4 for 4 h, 10
was produced. Integration of H NMR peaks versus an
1
internal standard revealed that the 10 produced came
from [7 + 8], and within experimental error, none came
from 9. More DDM was added to this sample, and it was
refluxed for an additional 3 h, which gave a product
mixture consisting of 9, 10, and products of self-reaction
of DDM, with [7 + 8] barely discernible by NMR.
Additionally, as a control, the ratio of [7 + 8] to 9 did
not change when fraction A was refluxed in CCl4 for 4 h
in the presence of Rh2(OAc)4 but without DDM. This
means that in principle one could increase the yield of
10 by converting the [7 + 8] present in fraction A to 10
by reacting fraction A with DDM and Rh2(OAc)4 in
refluxing CCl4, although we have not undertaken that
optimization.
The relatively high yield and easy isolation suggested
that 10 might be made in larger amounts. In fact, we
have scaled up the reaction, reducing the amount of
catalyst from 0.85 mol % to 0.45 mol % to lower the cost,
and have obtained over 1.5 g of 10 in a single run (44%
yield).
Tricyclic structures such as 10 have been termed “bis-
σ-homobenzenes” and have been synthesized by a variety
of routes.17 These syntheses are multistep and afford the
bis-σ-homoaromatic in low overall yield. Direct, one-pot
double cyclopropanation of benzene usually fails, as
discussed. We believe it is accurate to state that the
reaction of eq 3 represents the most convenient and most
efficient synthesis to date of any bis-σ-homoaromatic.
The Rh2(OAc)4/DDM methodology was investigated
with other aromatic substrates. Naphthalene (liquid, as
solvent) gave 12 in 31% yield. Anthracene (in CCl4, DDM/
anthracene ) 1.8:1) gave 16 in 20% yield, based on
anthracene reacted. Toluene led to an intractable mixture
of isomeric products. Linstrumelle, et al. found thermal
and photochemical Bu¨chner reactions of DDM with
toluene also gave mixtures of isomers.7 Anisole gave 17
cleanly in 70% isolated yield, with no sign of double
cyclopropanation, which is reminiscent of the behavior
of 2-diazo-1,3-indanedione with anisole.18
It is reasonable to state that, mechanistically, diadduct
10 results from a competition between norcaradiene 8
(13) Huisgen, R.; J uppe, G. Liebigs Ann. Chem. 1961, 646, 1-9.
(14) Tanny, S. R.; Grossman, J .; Fowler, F. W. J . Am. Chem. Soc.
1972, 94, 6495-6501.
(15) Davies, H. M. L.; Young, W. B.; Smith, H. D. Tetrahedron Lett.
1989, 30, 4653-4656.
(16) Domareva-Mandelshtam, T. V.; Dyakonov, I. A. Zh. Obshch.
Khim. 1964, 34, 3844-3845.
The yield of 10 may be increased, if desired. When
fraction A, viz. a mixture of [7 + 8] and 9, was treated
(8) (a) Weygand, F.; Schwenke, W.; Bestmann, H. J . Angew. Chem.
1958, 70, 506. (b) Reference 7. (c) J ones, M., J r.; Ando, W.; Hendrick,
M. E.; Kulczycke, A., J r.; Howley, P. M.; Hummel, K. F.; Malament,
D. S. J . Am. Chem. Soc. 1972, 94, 7469-7479. (d) In one instance,
these compounds were not found as products of thermal or photo-
chemical reactions of DDM with benzene: Ciganek, E. J . Org. Chem.
1965, 30, 4366-4367.
(17) (a) Prinzbach, H.; Stusche, D. Angew. Chem., Int. Ed. Engl.
1970, 9, 799-800. (b) Whitlock, H. W., J r.; Schatz, P. F. J . Am. Chem.
Soc. 1971, 93, 3837-3839. (c) Engelhard, M.; Lu¨ttke, W. Angew. Chem.,
Int. Ed. Engl. 1972, 11, 310. (d) Taylor, R. T.; Paquette, L. A. Angew.
Chem., Int. Ed. Engl. 1975, 14, 496. (e) Kaupp, G.; Ro¨sch, K. Angew.
Chem., Int. Ed. Engl. 1976, 15, 163-164. (f) Paquette, L. A.; Taylor,
R. T. J . Am. Chem. Soc. 1977, 99, 5708-5715. (g) Spielmann, W.;
Kaufmann, D. de Meijere, A. Angew. Chem., Int. Ed. Engl. 1978, 17,
440-441. (h) Christl, M.; Lechner, M. Chem. Ber. 1982, 115, 1-13. (i)
Kaufmann, D.; Fick, H.-H.; Schallner, O.; Spielmann, W.; Meyer, L.-
U.; Go¨litz, P.; de Meijere, A. Chem. Ber. 1983, 116, 587-609.
(18) Rosenfeld, M. J .; Ravi Shankar, B. K.; Shechter, H. J . Org.
Chem. 1988, 53, 2699-2705.
(9) Dalrymple, D. L.; Taylor, S. P. B. J . Am. Chem. Soc. 1971, 93,
7098.
(10) Pomerantz, M.; Levanon, M. Tetrahedron Lett. 1991, 32, 995-
998.
(11) Huisgen, R.; J uppe, G. Chem. Ber. 1961, 94, 2332-2349.
(12) Huisgen, R. J uppe, G. Tetrahedron 1961, 15, 7-17.

Notes
J . Org. Chem., Vol. 66, No. 14, 2001 4947
and benzene for available Rh-carbenoid. The yield of 10
is startling when one considers that 8 is at an enormous
statistical disadvantage versus benzene, the solvent. One
way to explain the atypically high yield of 10 is to posit
that 8, as a diene, is simply so much more reactive than
benzene that it may compensate for its statistical disad-
vantage. To test this idea qualitatively, an equimolar
mixture of benzene and 1,3-cyclohexadiene (a model for
8) was reacted with a deficiency of DDM and Rh2(OAc)4.
Within the limits of NMR detection, no products of
reaction of DDM with benzene were found. Therefore, the
diene is probably at least 100 times more reactive than
benzene, and this increase in reactivity may also apply
to 8 vs benzene. However, using reasonable assumptions,
we estimate a reactivity ratio of ca. 550 would be required
to account for the observed yield of 10. It is interesting
to note that although the use of Rh2(O2CCF3)4 in the
reaction of eq 3 had a large effect on the yield of 10,
repeating the benzene-1,3-cyclohexadiene competition
using Rh2(O2CCF3)4 had almost no effect on that outcome.
If 8 is 550 times more reactive than benzene toward
carbenoids, one may ask why 10 was not produced in
good yield in the many previously reported Bu¨chner
reactions of benzene. A plausible answer is that the much
more reactive carbenes used previously were unselective,
reacting statistically, leading to monocyclopropanation
products exclusively. In the present instance, the lower
reactivity of the DDM-Rh2(OAc)4 pair19 allowed a selec-
tive reaction to occur. According to the general mecha-
nism for carbenoid cyclopropanation proposed by Doyle
et al.,20 the Rh-carbenoid reversibly forms a π-complex
with the substrate, which was an alkene in the cases
treated originally by Doyle et al., but in the present case
is either benzene or diene 8 (or possibly triene 7). If the
equilibrium constant for π-complex formation were greater
for diene 8 than for benzene, this would constitute
another factor tending to favor reaction of the carbenoid
with 8 in preference to benzene. However, such equilib-
rium constants are not known. Although complexes of
rhodium(II) carboxylates with a wide variety of Lewis
bases have been studied,21 it would be inadvisable to
extrapolate those results to the question of complexes of
Rh(O2CCH3)4RhdC(COOCH3)2. A less attractive ratio-
nale for the high yield of 10 is to suggest 8 is efficiently
trapped by DDM by a 1,3-dipolar cycloaddition to form
18,22 which subsequently loses N2 and rearranges to 10.
However, we have no evidence for the existence of 18.
Compound 10 is thermally stable, having been recov-
ered unchanged after heating at 200-260 °C for 40 min.
The ready availability of 10 herein reported would
obviously be of greater importance if 10 could be func-
tionalized easily and selectively. As one step in that
direction, we have found mild hydrolysis cleaved two of
the four ester groups of 10, affording exclusively the exo,-
exo diacid 19 in 83% yield (eq 4). Proof of the stereo-
chemistry of 19 was nontrivial and required X-ray
crystallography (see the Supporting Information), con-
firming, as expected, that only the sterically more acces-
sible exo ester groups of 10 were hydrolyzed, and revealed
that 19 possesses crystallographic C2 symmetry. Dif-
ferential scanning calorimetry (DSC) of 19 up to 500 °C
revealed no processes other than loss of water of hydra-
tion and melting; i.e., decarboxylation apparently does
not occur within the time frame of the DSC experiment.
Also, tetraester 10 may be reduced nonselectively to
tetraol 20 (eq 5).
Only one other nonheteroatom σ-homobenzene has
been examined crystallographically: 21.23,24 As far as we
are aware, the present crystallographic results are the first
for bis-σ-homobenzenes. At first glance, 21 would appear
to be an interesting compound with which to compare
the structures of 10 and 19. However, the presence in
21 of neighboring cis cyclopropanes bearing geminal
methyls causes distortions not present in 10 and 19. For
example, the central six-membered ring of 21 is bent,
boatlike, at a dihedral angle of 163°. But, in 10 and 19,
the six carbons derived from benzene are planar, despite
this ring’s containing four formally tetrahedral carbons.
The root mean square (rms) deviation of these atoms from
the best plane through them is 0.004 Å, which is in the
neighborhood of the esd’s of the fractional coordinates of
the atoms. The six endocyclic C-C-C angles in this
“planar cyclohexene” average 120.0 ( 2.1° for 10 and 19.
Heteroatom-containing tris-σ-homoaromatics have “pla-
nar cyclohexane” rings.24 These curious planar entities
have been discussed recently by Dodziuk.25 The cyclo-
(19) Pirrung, M. C.; Morehead, A. T., J r. J . Am. Chem. Soc. 1996,
118, 8162-8163.
(20) Doyle, M. P. Chem. Rev. 1986, 86, 919-939.
(21) (a) Doyle, M. P.; Coleman, M. R.; Chinn, M. S. Inorg. Chem.
1984, 23, 3684-3685. (b) Drago, R. S.; Tanner, S. P.; Richman, R. M.;
Long, J . R. J . Am. Chem. Soc. 1979, 101, 2897-2903. (c) Doyle, M. P.;
Wang, L. C.; Loh, K.-L. Tetrahedron Lett. 1984, 25, 4087-4090. (d)
Cotton, F. A.; Dikarev, E. V.; Petrukhina, M. A.; Stiriba, S.-E.
Polyhedron 2000, 19, 1829-1835 (e) Pirrung, M. C.; Morehead, A. T.,
J r. J . Am. Chem. Soc. 1994, 116, 8991-9000. (f) Doyle, M. P.;
Mahapatro, S. N.; Caughey, A. C.; Chinn, M. S.; Colsman, M. R.; Harn,
N. K.; Redwine, A. E. Inorg. Chem. 1987, 26, 3070-3072. (g) Bilgrien,
C.; Drago, R. S.; Vogel, G. C.; Stahlbush, J . Inorg. Chem. 1986, 25,
2864-2866. (h) Schurig, V. Inorg. Chem. 1986, 25, 945-949. (i)
Chavan, M. Y.; Lin, X. Q.; Ahsan, M. Q.; Bernal, I.; Bear, J . L.; Kadish,
K. M. Inorg. Chem. 1986, 25, 1281-1288. (j) Drago, R. S.; Telser, J .
Inorg. Chem. 1984, 23, 2599-2606. (k) Boyar, E. B.; Robinson, S. D.
J . Chem. Soc., Dalton Trans. 1985, 629-633.
(23) Kru¨ger, C.; Roberts, P. J . Cryst. Struct. Commun. 1974, 3, 459-
462.
(24) X-ray crystallographic studies of several heteroatom-containing
homoaromatics (e.g., epoxides or aziridines) have been reported: (a)
Ba´nhidai, B.; Scho¨llkopf, U. Angew. Chem., Int. Ed. Engl. 1973, 12,
836. (b) Littke, W.; Dru¨ck, U. Angew. Chem., Int. Ed. Engl. 1974, 13,
539-540. (c) Vogel, E.; Breuer, A.; Sommerfeld, C.-D.; Davis, R. E.;
Liu, L.-K. Angew. Chem., Int. Ed. Engl. 1977, 16, 169-170. (d)
Schwesinger, R.; Piontek, K.; Littke, W.; Prinzbach, H. Angew. Chem.,
Int. Ed. Engl. 1985, 24, 318-319.
(22) Atasoy, B.; Balci, M.; Bu¨yu¨kgu¨ngo¨r, O. Tetrahedron Lett. 1987,
28, 555-558.
(25) Dodziuk, H. Top. Stereochem. 1994, 21, 351-380.

4948 J . Org. Chem., Vol. 66, No. 14, 2001
Notes
of benzene, and 7 mg (0.01 mmol) of Rh2(O2CCF3)4 was added.
The mixture was refluxed until TLC showed no DDM present,
about 1 h. Evaporation of the solvent and NMR analysis of the
residue indicated the following products: [7 + 8], 64%; 9, 32%,
10, 4%.
propanes of 10 and 19 are better described as isosceles
than equilateral triangles. The average length of the
“nonfused” cyclopropane C-C bonds in 10 and 19, 1.530
Å, is in the uppermost quartile of the sample of 888
cyclopropane C-C bond lengths surveyed by Allen et al.26
The remaining cyclopropane C-C bonds, those common
to the cyclohexene ring, average 1.491 Å, which is in the
lowest quartile of the same dataset. Similar values are
observed for cyclopropane-1,1-dicarboxylic acid.27 The
angles between the plane of the cyclopropane ring and
the best plane through the planar cyclohexene ring are
109.0(1)°and 110.5(1)° for 10 and 109.2(1)° for 19. In 21,
the unique cyclopropane makes a dihedral angle of 110.1°
with its half of the six-membered ring.23 Because 19
crystallized as a dihydrate, the usual carboxylic acid
dimer H-bonds28 were replaced by a H-bond network
including waters of crystallization.
Ben zo-5,5,8,8-tetr a ca r bom eth oxytr icyclo[5.1.0.04,6]oct-2-
en e, 12. A mixture of 6.891 g (53.76 mmol) naphthalene and
0.413 g (2.61 mmol) dimethyl diazomalonate was heated in an
86 °C oil bath until the mixture had liquefied. To this was added
11 mg (0.025 mmol) Rh2(OAc)4. The reaction was heated and
stirred for 5 h, at which time TLC indicated no dimethyl
diazomalonate remained. Excess naphthalene was sublimed, and
the residue was subjected to silica gel chromatography using
1:3 (v/v) EtOAc/hexanes as eluent. Two fractions were col-
lected: A, 176 mg; B, 155 mg. Fraction A was a mixture of benzo-
[b]-7,7-dicarbomethoxybicyclo[4.1.0]hept-2-ene, the monocyclo-
propanation product analogous to [7 + 8] that was identified by
comparison with published spectra,10 dimethyl 1-naphthylma-
lonate, and dimethyl 2-naphthylmalonate. Fraction B was
1
diadduct 12: 30.6% yield; mp 159-60 °C; H NMR (250 MHz,
The availability on a larger scale of thermally stable,
C2-symmetric bis-σ-homoaromatics 10, 12, and 16 in one
easy step and 19 and 20 in two easy steps will stimulate
a variety of imaginative uses of these intriguing com-
pounds, we suspect.
CDCl3) δ 7.24-7.15 (m, 4H), 3.77 (s, 6H), 3.42 (s, 6H), 2.73
(appar. AB quartet, J ) 9.8 Hz, 4H); 13C NMR (62.9 MHz, CDCl3)
δ 169.0 (quat), 165.8 (quat), 130.5, 128.6 (quat), 127.7, 52.9, 41.8
(quat), 29.9, 26.7.
Na p h th o(b)-5,5,8,8-tetr a ca r bom eth oxytr icyclo[5.1.0.04,6]-
oct-2-en e, 16. A suspension of 1.12 g (6.29 mmol) of anthracene
in 30 mL of CCl4 was heated at reflux until the anthracene
dissolved. To this solution was added 12 mg (0.027 mmol) of Rh2-
(OAc)4 and 1.78 g (11.3 mmol) of dimethyl diazomalonate. The
reaction mixture was refluxed for 5.5 h, at which time TLC
indicated no diazomalonate remaining. Solvent was removed at
the rotary evaporator, and the residue was purified by silica gel
column chromatography, using a step-gradient from pure hex-
anes to 1:2 (v/v) EtOAc/hexanes, in which anthracene (0.613 g)
was recovered. Two other fractions were collected: A, 95 mg and
B, 261 mg. Fraction A was naphtho[b]-7,7-dicarbomethoxybicyclo-
[4.1.0]hept-2-ene, the monocyclopropanation product analogous
to [7 + 8]: 8.3% yield; 1H NMR (250 MHz, CDCl3) δ 7.93 (s,
1H), 7.82-7.76 (m, 2H), 7.56 (s, 1H), 7.45-7.41 (m, 2H), 6.67
(d, J ) 9.8 Hz, 1H), 6.25 (m, 1H), 3.82 (s, 3H), 3.25 (d, J ) 8.5
Hz, 1H), 3.33 (s, 3H), 3.02 (m, 1H); 13C NMR (62.9 MHz, CDCl3)
δ 171.8, 165.3, 133.5, 133.1, 129.6, 129.1, 128.4, 128.1, 127.6,
126.6, 126.3, 126.2, 123.3, 53.4, 52.4, 34.5, 33.8, 31.0. Fraction
B was diadduct 16: 20.8% yield; mp 229-230 °C; 1H NMR (250
MHz, CDCl3) δ 5.04. 7.74-7.70 (m, 4H), 7.43-7.39 (m, 2H), 3.79
(s, 6H), 3.36 (s, 6H), 2.85 (AB quartet, J ) 9.5 Hz, 4H); 13C NMR
(62.9 MHz, CDCl3) δ 169.2, 166.0, 133.1, 129.9, 127.6, 126.7,
126.3, 53.2, 52.6, 41.8, 30.8, 26.8. Anal. Calcd for C24H22O8: C,
65.75; H, 5.06. Found: C, 65.81; H, 5.04. Both yields take into
account recovered anthracene.
Exp er im en ta l Section
5,5,8,8-Tetr acar bom eth oxytr icyclo[5.1.0.04,6]oct-2-en e, 10.
(a) To a solution of 335 mg (2.12 mmol) dimethyl diazomalonate
in 15 mL of benzene was added 8 mg (0.02 mmol) of Rh2(OAc)4,
and the resulting green mixture was brought to reflux. After 4
h, TLC indicated the diazo compound had been consumed, and
heating was stopped. After cooling and filtration, solvent was
removed at the rotary evaporator and the residue subjected to
column chromatography on silica gel (60-200 mesh; EtOAc/
hexanes 1:2 (v/v)). Fraction A (121 mg) was a mixture of [7 + 8]
and 9. For [7 + 8]: 1H NMR (250 MHz, CDCl3) δ 6.47 (m, 2H),
6.38 (m, 2H), 5.04 (dd, J ) 7.4, 0.8 Hz, 2H), 3.69 (s, 6H); 13C
NMR (63 MHz, CDCl3, proton chemical shift correlation from
2D H-C correlation experiment given in italics) δ 169.3 (quat),
128.9 (6.47), 126.2 (6.38), 100.04 (5.04), 51.4 (quat), 53.1 (3.69).
For 9: 1H NMR (250 MHz, CDCl3) δ 7.37 (m, 5H), 4.65 (s, 1H),
3.75 (s, 6H); 13C NMR (63 MHz, CDCl3, proton chemical shift
correlation from 2D H-C correlation experiment given in italics)
δ 168.7 (quat), 132.7 (quat), 129.3 (7.37), 128.8 (7.37), 128.4
(7.37), 57.7 (4.65), 52.9 (3.75).
Fraction B was an oil that later solidified to afford 208 mg
1
(58%) 10: mp 122-123 °C; H NMR (250 MHz, CDCl3) δ 5.71
Tr icyclo[5.1.0.04,6]oct-2-en e-5,5,8,8-tetr a ca r boxylic Acid
en d o,en d o-Dim eth yl Ester , 19. To a solution of 218 mg (0.644
mmol) of tetraester 10 in 11 mL of H2O/THF (4:7 (v/v)) was
added 95 mg (2.26 mmol) of LiOH‚H2O. The mixture was stirred
at room temperature for 2 h, and 3 M HCl was added to pH )
3. After the inside of the flask was scratched with a glass rod,
precipitation began and was allowed to continue for 0.5 h at ice-
bath temperature. The solid 19 was filtered and air-dried: 166
(m, 2H, (H2, H3)), 3.75 (s, 6H), 3.74 (s, 6H), 2.50 (d with
unresolved splitting, J ) 9.5 Hz, 2H, (H6, H7)), 2.02 (dd with
unresolved splitting, J ) 9.5, 3.1 Hz, 2H, (H1, H4)); 13C NMR
(63 MHz, CDCl3, proton chemical shift correlation from 2D H-C
correlation experiment given in italics) δ 169.2 (quat), 166.0
(quat), 122.8 (CH, 5.71), 52.9 (CH3, 3.74), 52.6 (CH3, 3.75), 42.3
(quat), 25.6 (CH, 2.50), 24.9 (CH, 2.02); EI-MS, m/e (rel intensity)
338 (M, 0.2), 320 (0.8), 306 (1.2), 274 (45), 247 (39), 219 (44),
215 (49), 208 (25), 187 (44), 170 (20), 129 (23), 105 (31), 89 (40),
59 (100); IR 1729, 1253 cm-1. Anal. Calcd for C16H18O8: C, 56.80;
H, 5.36. Found: C, 56.71; H 5.42. A crystal 0.80 × 0.40 × 0.20
mm was selected for X-ray crystallography with 0.71073 Å
radiation: monoclinic a ) 8.290(4) Å, b ) 21.670(8) Å, c )
10.051(5) Å, â ) 113.16(3)°; P21/n; Z ) 4; 3120 reflections were
collected, 2920 independent (Rint ) 0.0268), 0 e h e 9, -22 e k
e 25, -11 e l e 11. Full-matrix least-squares refinement on
F2, data-to-parameter ratio ) 13.3, gave goodness-of-fit 1.025,
R1 ) 0.0460, wR2 ) 0.1151 (I > 2σ(I)), R1 ) 0.0617, wR2 )
0.1912 (all data).
1
mg (83.0%); mp 201-202 °C; H NMR (250 MHz, acetone-d6) δ
6.3-4.9 (br s), 5.68 (app t, J ) 2.1 Hz, 2H) 3.68 (s, 6H), 2.42 (d,
J ) 9.3 Hz, 2H), 1.94 (br d, J ) 9.4 Hz, 2H); 13C NMR (63 MHz,
acetone-d6) δ 170.2, 166.8, 123.5, 52.6, 43.0, 26.2, 25.4. Anal.
Calcd for C14H14O8: C, 54.20; H, 4.55. Found: C, 54.10; H, 4.54.
A crystal 0.38 × 0.32 × 0.08 mm was selected for X-ray
crystallography with 0.71073 Å radiation: monoclinic a
)
15.133(7) Å, b ) 10.569(5) Å, c ) 10.436(4) Å, â ) 105.32(3)°;
C2/c; Z ) 4; 1482 reflections were collected, 1426 independent
(Rint ) 0.0132), 0 e h e 18, 0 e k e 12, -12 e l e 11. Full-
matrix least-squares refinement on F2, data-to-parameter ratio
) 11.7, gave goodness-of-fit 1.074, R1 ) 0.0425, wR2 ) 0.0922
(I > 2σ(I)), R1 ) 0.0670, wR2 ) 0.1038 (all data). The structure
revealed the compound had crystallized as a dihydrate, viz.
C14H14O8‚2H2O.
(b) The reaction was repeated with Rh2(O2CCF3)4 instead of
Rh2(OAc)4. DDM (315 mg, 1.99 mmol) was dissolved in 15 mL
(26) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen,
A. G.; Taylor, R. J . Chem. Soc., Perkin Trans. 2 1987, S1-S19.
(27) Meester, M. A. M.; Schenk, H.; MacGillavry, C. H. Acta
Crystallogr. 1971, B27, 630-634.
5,5,8,8-Tetr a k is(h yd r oxym eth yl)tr icyclo[5.1.0.04,6]oct-2-
en e, 20. To a mixture of 0.45 g LiAlH4 (12 mmol) in 7 mL of dry
THF was added a solution of 0.35 g 10 (1.0 mmol) in 5 mL of
dry THF dropwise under a N2 atmosphere. The reaction was
(28) Leiserowitz, L. Acta Crystallogr. 1976, B32, 775-802.

Notes
J . Org. Chem., Vol. 66, No. 14, 2001 4949
warmed to 60 °C for 12 h. The reaction was cooled, and to it
were added sequentially 0.45 mL of water, 0.45 mL of 15%
aqueous NaOH, and 1.30 mL of water. The mixture was filtered
and the filtrate evaporated to give a pale yellow oil. When
triturated with 1:8 MeOH/EtOAc, the oil solidified affording 0.17
g (73%) of 20 as a colorless solid: mp 159-161 °C; 1H NMR (400
MHz, DMSO-d6 + D2O) δ 5.60 (m, 2H) 3.70 (AB quartet, J )
11.2 Hz, ∆ν ) 15.4 Hz, 4H), 3.33 (AB quartet, J ) 11.2, ∆ν )
22.5 Hz, 4H), 1.29 (d, J ) 8.6 Hz, 2H), 1.02 (d with unresolved
splitting, J ) 8.6 Hz, 2H); 13C NMR (100 MHz, DMSO-d6): 123.3,
65.1, 57.7, 36.2, 19.7, 18.2. Anal. Calcd for C12H18O4: C, 63.70;
H,8.02. Found: C, 63.43; H, 8.17.
Com p etition Exp er im en ts. (a) A mixture of 5.0 mL of
benzene (56 mmol), 5.0 mL of 1,3-cyclohexadiene (53 mmol), 0.69
g of DDM (4.4 mmol), and 12 mg of Rh2(OAc)4 (0.027 mmol) was
refluxed until TLC showed no DDM remaining, about 1.5 h.
NMR analysis of the residue after evaporation of solvents
showed no products of reaction of DDM with benzene (7-10).
The only product formed (aside from minute amounts of uni-
dentified sideproducts) was 7,7-dicarbomethoxybicyclo[4.1.0]-
hept-2-ene: 1H NMR (250 MHz, CDCl3) δ 5.93-5.67 (m, 2H),
3.72 (s, 3H), 3.71 (s, 3H), 2.21-1.59 (m, 6H); 13C NMR (63 MHz,
CDCl3) δ 170.1, 167.7, 127.8, 121.8, 52.6, 52.5, 40.8, 26.6, 24.5,
20.5, 16.0.
mmol), and the resultant mixture was heated at 80 °C for 3 h.
After evaporation of solvent, silica gel column chromatography
(THF/hexane 1:2 (v/v)) gave three fractions. However, NMR
spectroscopy revealed each fraction to be composed of a mixture
of several (most likely isomeric) compounds.
(b) An isole. To a mixture of 0.51 g (3.2 mmol) of DDM and
15 mL of anisole was added 10 mg of Rh2(OAc)4 (0.023 mmol),
and the resultant mixture heated to 80 °C for 4 h. At this time,
TLC showed no DDM present. Excess anisole was removed by
vacuum distillation and the residue purified by column chro-
matography on silica gel (EtOAc/hexanes 1:5(v/v)), to afford 0.53
g (70%) of dimethyl p-methoxyphenylmalonate: mp 44-45 °C
(lit.29 mp 44-47 °C); 1H NMR (250 MHz, CDCl3) δ 7.35-7.31
(m, 2H), 6.92-6.88 (m, 2H), 4.61 (s, 1H), 3.80 (s, 3H), 3.75 (s,
6H); 13C NMR (63 MHz, CDCl3) δ 169.0, 159.7, 130.5, 124.8,
114.3, 56.9, 55.4, 52.9.
Ack n ow led gm en t. The authors gratefully acknowl-
edge the support of the Auburn University Department
of Chemistry. We warmly thank Professor R. A. Don-
nelly for fruitful discussions.
Su p p or tin g In for m a tion Ava ila ble: Crystallographic
data for 10 and 19‚2H2O and 1H and 13C NMR spectra of
fraction A of the reaction that led to 12 and of the reaction
that led to 16. This material is available free of charge via
the Internet at http://pubs.acs.org.
(b) The benzene/cyclohexadiene competition described above
was repeated using Rh2(O2CCF3)4 instead of Rh2(OAc)4 (1.93 g
of 1,3-cyclohexadiene (24.1 mmol), 1.88 g of benzene (24.1 mmol),
0.44 g of DDM (2.8 mmol), and 7 mg of Rh2(O2CCF3)4 (0.01
mmol)). After 1.5 h reflux, DDM was consumed. NMR of the
residue after removal of volatiles was almost identical to that
obtained in the Rh2(OAc)4 competition, the only difference being
a slightly higher level of unidentified side products.
J O010277X
(29) (a) Dannhardt, G.; Meindl, W.; Schober, B. D.; Kappe, T. Eur.
J . Med. Chem. 1991, 26, 599-604 (b) Baciocchi, E.; Dell’Aira, D.;
Ruzziconi, R. Tetrahedron Lett. 1986, 27, 2763-2766 (c) Gillespie, R.
J .; Porter, A. E. A. J . Chem. Soc., Chem. Commun. 1979, 50-51.
R ea ct ion of Dim et h yl Dia zom a lon a t e w it h Ben zen e
Der iva tives. (a ) Tolu en e. To a mixture of 0.99 g of DDM (6.3
mmol) in 15 mL of toluene was added 13 mg of Rh2(OAc)4 (0.029