Favorskii reactions of a bromoquadricyclanone

Full text of DOI:10.1021/jo951672o

J . Org. Chem. 1996, 61, 7611-7613
7611
Ta ble 1. Geom etr y Op tim ized HF /6-31G* Tota l En er gies,
Hea ts of F or m a tion (∆Hf), a n d Str a in En er gies (SE) for
Qu a d r icycla n e, P r ism a n e, Hom ocu ba n e, Cu ba n e,
Hom op en ta p r ism a n e, a n d P en ta p r ism a n e
F a vor sk ii Rea ction s of a
Br om oqu a d r icycla n on e
Lisa E. Boyer, J ason Brazzillo, Mark A. Forman,* and
Brian Zanoni
∆Hf
SE
compound
E (HF) au
(kcal/mol)a (kcal/mol)b
quadricyclane
prismane
homocubane
cubane
homopentaprismane -423.4474219
pentaprismane
-269.61822
80.9d
136.2
93.7
97.2
147.6
113.9
161.5
95.0
Department of Chemistry, Saint J oseph’s University,
Philadelphia, Pennsylvania 19131
-230.5032877
-346.5044871
-307.3939067
146.3c
71.0
Received September 12, 1995
-384.3418194
120.5
139.5
The [n]-prismanes are an interesting class of highly
symmetrical (CH)n polyhedranes whose novel structure
and reactivity has stimulated significant synthetic1 and
theoretical2 attention. Prismane ([3]-prismane) (1),3 cu-
bane ([4]-prismane) (2),4 and pentaprismane ([5]-pris-
mane) (3)5 have been synthesized and are known to have
D3h, Oh, and D5h symmetry, respectively. Whereas pris-
mane (1) was prepared via photochemical denitrogena-
tion, both the cubane and pentaprismane skeletons were
constructed using the Favorskii ring-contraction. More-
over, most of the synthetic approaches toward the
unknown higher order prismanes (4-6)6 have also uti-
lized the Favorskii ring-contraction and with good reason.
This reaction has proven to be a very reliable method
for the synthesis of not only prismatic ring systems, but
also of a variety of other strained ring compounds. As
yet, however, there have been no reports of the use of
this reaction for the synthesis of the [3]-prismane skel-
eton. This is likely a result of the fact that Favorskii
ring-contractions are typically run at elevated tempera-
tures and that [3]-prismanes are thermally labile. In-
deed, the parent hydrocarbon 1 has a half life of 11 h at
90 °C. Nonetheless, we were intrigued by this approach,
particularly because the immediate product of Favorskii
ring-contraction, [3]-prismane carboxylic acid (11), could
then be metalated using the protocol of Eaton.7 It would
then be possible to synthesize a host of new substituted
[3]-prismanes. Finally, the fact that a readily available
Favorskii precursor quickly unveiled itself further en-
hanced the appeal of this approach.
a
Calculated using the group equivalent method of Ibrahim and
Schleyer (Ibrahim, M. R.; Schleyer, P. J . Comput. Chem. 1985, 6,
157. Calculated from Benson group equivalents (Benson, S. W.;
Cruickshank, F. R.; Golden, D. M.; Haugen, G. R.; O’Neal, H. E.;
Rodgers, A. S.; Shaw, R.; Walsh, R. Chem. Rev. 1969, 69, 279.).
c The experimental heat of formation is 148.7 kcal/mol (Wiberg,
b
d
K. B. Angew. Chem., Int. Ed. Engl. 1986, 25, 312.). The experi-
mental heat of formation is 79.5 kcal/mol (Rodgers, D. W.; Ling,
S. C.; Girellini, R. S.; Holmes, T. J .; Allinger, N. L. J . Phys. Chem.
1980, 84, 1810.).
quadricyclane to [3]-prismane was computed using ge-
ometry optimized ab initio calculations at the HF/6-31G*
level, and this value was compared to the strain energy
increase for the transformation of homocubane to cubane
and homopentaprismane to pentaprismane. The results
of these calculations are shown in Table 1. It is note-
worthy that the increase in strain energy of 50.4 kcal/
mol for the proposed quadricyclane to prismane trans-
formation is not significantly different than that for the
homocubane to cubane (47.6 kcal/mol) and homopenta-
prismane to pentaprismane (44.5 kcal/mol) transforma-
tions. Indeed this value appears to fall below the
threshold proposed by Mehta and Osawa,8 further bol-
stering our hope that the Favorskii ring-contraction to
the [3]-prismane skeleton would be successful.
Our synthesis of an appropriate [3]-prismane Favorskii
precursor relied on the work of Klumpp and co-workers
who have shown that a variety of cyclopropyl carbinols,
including quadricyclanol (7), may be readily lithiated via
treatment with an alkyllithium.9 Expanding on this
work of Klumpp, Szeimies reported that lithiation of
quadricyclanol (7) with 2 equiv of n-butyllithium, followed
by addition of dibromoethane, afforded an unseparated
85:15 mixture of brominated alcohols 8 and 9.10 We
found that 8 and 9 could be separated via careful silica
gel chromatography, provided that the eluent contains
2% triethylamine to prevent isomerization of the quad-
ricyclanol to the norbornadiene. Oxidation of pure bro-
mohydrin 8 was then carried out with tetrapropyl-
ammonium perruthenate(VII) (TPAP)/N-methylmorpho-
In order to evaluate the feasibility of this approach,
the increase in strain energy for the conversion of
(6) (a) Eaton, P. E.; Chakraborty, U. R. J . Am. Chem. Soc. 1978,
100, 3634. (b) Mehta, G.; Padma, S. Tetrahedron 1991, 47, 7783. (c)
Mehta, G.; Padma, S. Tetrahedron 1991, 47, 7807. (d) Forman, M. A.;
Dailey, W. P. J . Org. Chem. 1993, 58, 1501. (d) Mehta, G.; Reddy, S.
H. K.; Padma, S. Tetrahedron 1991, 47, 7821.
(7) (a) Eaton, P. E.; Castaldi, G. J . Am. Chem. Soc. 1985, 107, 724.
(b) Eaton, P. E.; Higuchi, H.; Millikan, R. Tetrahedron Lett. 1987, 28,
1055. (c) Eaton, P. E.; Cunkle, G.; Marchioro, G.; Martin, R. M. J . Am.
Chem. Soc. 1987, 109, 948. (d) Eaton, P. E.; Martin, R. M. J . Org. Chem.
1988, 53, 2728. (e) Eaton, P. E.; Daniels, R. G.; Casucci, D. ; Cunkle,
G. T.; Engel, P. J . Org. Chem. 1987, 52, 2100.
(8) Osawa, E.; Barbiric, D. A.; Lee, O. S.; Kitano, Y.; Padma, S.;
Mehta, G. J . Chem. Soc., Perkin Trans. 2 1989, 1161.
(9) (a) Klumpp, G. W.; Kool, M.; Schakel, M.; Schmitz, R. F.;
Boutkan, C. J . Am. Chem. Soc. 1979, 101, 7065. (b) Klumpp, G. W.;
Kool, M.; Veefkind, A. H.; Schakel, M.; Schmitz, R. F. Recl. Trav. Chim.
Pays-Bas 1983, 102, 542. (c) Klumpp, G. W. Recl. Trav. Chim. Pays-
Bas 1986, 105, 1.
(1) For three recent [n]-prismane reviews see (a) Forman, M. A.;
Dailey, W. P. Org. Prep. Proc. Int. 1994, 26, 291. (b) Mehta, G.; Padma,
S. In Carbocyclic Cage Compounds; Osawa, E., Yonemitsu, O., Eds.,
VCH Publishers Inc.: New York, 1992; p 183. (c) Mehta, G. In Strain
and Its Implications in Organic Chemistry; de Meijere, A., Blechert,
S., Eds., NATO ASI Series, Kluwer Academic Publishers: Dordrecht,
1989, 273, 269.
(2) Disch, R. L.; Schulman, J . M. J . Am. Chem. Soc. 1988, 110, 2102.
(b) Dailey, W. P. Tetrahedron Lett. 1987, 28, 5787. (c) Mehta, G.;
Padma, S.; Osawa, E.; Barbiric, D. A.; Mochizuki, Y. Tetrahedron Lett.
1987, 28, 1295.
(3) Katz T. J .; Acton, N. J . J . Am. Chem. Soc. 1973, 95, 2738.
(4) Eaton, P. E.; Cole, T. W. J . Am. Chem. Soc. 1964, 86, 3157.
(5) (a) Eaton, P. E.; Or, Y. S.; Branca, S. J . J . Am. Chem. Soc. 1981,
103, 2134. (b) Eaton, P. E.; Or, Y. S.; Branca, S. J .; Shankar, B. K. R.
Tetrahedron 1986, 42, 1621.
(10) Heywang, U.; Szeimies, G. Chem. Ber. 1990, 123, 121.
S0022-3263(95)01672-0 CCC: $12.00 © 1996 American Chemical Society

7612 J . Org. Chem., Vol. 61, No. 21, 1996
Notes
Sch em e 1
which, in water, takes place in preference to hydroxide
attack at the carbonyl. This ionization is further pro-
moted when lithium hydroxide is used as the base, for
10 reacts faster and at lower temperature in aqueous
LiOH (entry 3) than in aqueous KOH (entry 4). Indeed,
it is well-known that lithium ion can act as a Lewis acid,11
and here it appears to enhance C-Br ionization and
subsequent decomposition of 10. In further support of
this theory, stirring an aqueous solution of 10 in LiClO4
at 80 °C for 4 h leads to complete destruction of the
starting bromo ketone 10 and recovery of little organic
solvent soluble material. In addition, refluxing an aque-
ous solution of 10 for 17 h leads to complete destruction
of this bromo ketone. Furthermore, heating 10 in acid-
free deuterochloroform for 17 h leads to little loss of
starting material. No bromobenzene was detected in
either experiment. Finally, attempts at running these
reactions for shorter time periods led only to partial
recovery of starting bromo ketone 10; the amount of 10
recovered was inversely proportional to reaction time.
In summary, attempts to prepare the [3]-prismane
skeleton via Favorskii ring-contraction under a variety
of conditions have not been successful. In nonaqueous
solvents, only benzoic acid is obtained. Given the ther-
mal instability of [3]-prismanes, this result is not sur-
prising, though we cannot eliminate the possibility that
a Haller-Bauer cleavage pathway to the benzoic acid is
operative. Reactions run in aqueous base led to prefer-
ential C-Br bond ionization and ultimate destruction of
the starting bromo ketone 10 prior to hydroxide attack
at the ketone. Alternative approaches to [3]-prismane
carboxylic acid, and ultimately the parent hydrocarbon,
are underway in these laboratories.
Sch em e 2
Ta ble 2. Su m m a r y of Attem p ted F a vor sk ii
Rin g-Con tr a ction s of Br om o Keton e 10
reaction conditions
product
yield
NaOH, THF, reflux, 4 d
NaOH, toluene, reflux, 12 h
LiOH, H2O, 80 °C, 8 h
KOH, H2O, reflux, 26 h
benzoic acid
benzoic acid
decomposition
decomposition
96%
98%
-
Exp er im en ta l Section
Gen er a l Meth od s. All solvents and reagents were reagent
grade or better and were used without purification unless
otherwise indicated. Prior to drying at 125 °C for 5-6 h, all
glassware was washed sequentially with concentrated NH4OH,
5% aqueous EDTA, and distilled water to prevent isomerization
of the quadricyclane ring system. 1H NMR spectra were
obtained in CDCl3 at 500 MHz with CHCl3 (δ 7.24) as an internal
standard. 13C NMR spectra were obtained at 125 MHz with
CDCl3 (δ 77.0) as an internal standard. Reactions were moni-
tored by thin-layer chromatography and were carried out on
silica gel plates using anisaldehyde as a developing stain. Flash
chromatography was carried out using the method of Still.12
Melting points are reported uncorrected.
-
line N-oxide (NMO) in 97% yield to afford bromo ketone
10, ideally functionalized for Favorskii ring-contraction
(Scheme 1).
Because Favorskii ring-contractions are often sensitive
to reaction conditions, ring-contractions of 10 were car-
ried out under several different conditions as shown in
Table 2. Reaction with sodium hydroxide in refluxing
THF and reaction with sodium hydroxide in refluxing
toluene both furnished excellent yields of benzoic acid as
the sole product (entries 1 and 2). The benzoic acid may
be derived from the thermal isomerization of the putative
prismane carboxylate salt 12, or alternatively, from an
intermediate anion (13) resulting from Haller-Bauer
type cleavage of bromo ketone 10 (Scheme 2). However,
attempts at running these reactions at lower tempera-
tures in order to limit this potential isomerization were
met with failure, for only unreacted bromo ketone 10 was
recovered.
In contrast to the results found in nonaqueous solvents,
Favorskii ring-contractions carried out in aqueous base
(entries 3 and 4) took a very different course. In these
instances, only trace amounts of organic solvent soluble
material could be recovered. To eliminate the possibility
that the Haller-Bauer intermediate 13 had fragmented
to benzene as shown in Scheme 2, the reactions in
aqueous media were conducted in sealed tubes. However,
upon completion of these reactions, no benzene was
detected. Given these results, we attribute this difference
in reactivity in aqueous solvents to C-Br bond ionization
2-Br om otetr a cyclo[3.2.0.02,7.04,6]h ep ta n -3-ol (8). A crude
mixture of 2-bromotetracyclo[3.2.0.02,7.04,6]heptan-3-ol (8) and
1-bromotetracyclo[3.2.0.02,7.04,6]heptan-syn-3-ol (9) was prepared
as described by Szeimies and Heywang.10 The crude product
mixture was distilled at 51-55 °C/0.2 mm and the distillate
subjected to flash chromatography using 65% diethyl ether/33%
petroleum ether/ 2% triethylamine to afford pure 8 (Rf ) 0.5)
as a colorless oil whose spectral properties were identical to those
reported.
2-Br om otetr a cyclo[3.2.0.02,7.04,6]h ep ta n -3-on e (10). To a
stirred mixture of 514 mg (2.75 mmol) of 8, 2.49 g of crushed
4A molecular sieves, 850 mg (7.38 mmol) of 4-methylmorpholine
N-oxide in 10 mL of acetonitrile under nitrogen was added 90
mg (9 mol %) of TPAP all at once. After 17 min, TLC (65%
diethyl ether/35% petroleum ether, alcohol Rf ) 0.5; ketone Rf
) 0.7) indicated that the reaction was finished. The greenish
mixture was then placed on the rotovap to remove the aceto-
nitrile, and the resultant green-black solid was filtered through
a small pad of silica gel using 96% ethyl acetate/4% triethyl-
(11) (a) Pushin, A. N.; Tkachenko, S. E.; Martynov, I. V. Dokl. Akad.
Nauk SSSR 1988, 299, 154. (b) Forman, M. A.; Dailey, W. P. J .
Am.Chem. Soc. 1991, 113, 2761.
(12) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.

Notes
J . Org. Chem., Vol. 61, No. 21, 1996 7613
amine. The silica gel had previously been deactivated by
washing with the aforementioned solvent. Concentration of the
filtrate yielded 491 mg (97%) of slightly off-white crystalline 10
that was >95% pure by NMR. An analytical sample was
obtained by slow evaporation from hexane/ethyl acetate to yield
colorless crystals: mp 45-47 °C; IR (CDCl3, cm-1) 3060, 1760,
3. KOH/H2O. A magnetically stirred mixture of bromo ketone
10 (340 mg, 1.84 mmol) in 20% KOH (10 mL) was refluxed for
26 h at which time TLC indicated complete loss of 10. The dark
viscous mixture was cooled in an ice bath and acidified with 10%
HCl to ca. pH 3-4. The mixture was then extracted with 3:1
CH2Cl2/t-BuOH (4 × 35 mL). Combined organics were brine
washed (1 × 30 mL), dried over MgSO4, and concentrated to a
black oil. NMR revealed only two, broad unresolved peaks at δ
0.8-1.1.
1
1270; H NMR δ 2.47-2.49 (m, 2H), 2.36-2.38 (m, 2H), 1.24-
1.28 (m, 1H); 13C NMR δ 209.1, 30.7, 28.1, 22.3, 17.9; HRMS
(M+) calcd 184.9602, obsd 184.9600. Anal. Calcd for C7H5BrO:
C, 45.44; H, 2.72. Found C, 45.36; H, 2.74
4. LiOH/H2O. A magnetically stirred mixture of bromo
ketone 10 (198 mg, 1.07) in 1M LiOH (5.6 mL) was heated to 80
°C for 8 h at which time tlc indicated complete loss of 10. The
dark viscous mixture was cooled in an ice bath and acidified
with 10% HCl to ca. pH 3-4. The mixture was then extracted
with CH2Cl2 (4 × 25 mL). Combined organics were brine washed
(1 × 30 mL), dried over MgSO4, and concentrated to a small
amount (<20 mg) of yellowish oil. NMR revealed a broadened
baseline and two, larger, broad unresolved peaks at δ 0.8-1.1.
F a vor sk ii Rin g-Con tr a ction s. 1. NaOH/Toluene. A sus-
pension of 1.5 g of powdered NaOH in toluene (10 mL) was
refluxed for 45 min to azeotropically remove traces of water via
a Dean-Stark trap. The mixture was cooled to ambient tem-
perature and bromo ketone 10 (366 mg, 1.98 mmol) in toluene
(4 mL) was added and the mixture then refluxed for 12 h. After
12 h, TLC indicated that 10 had undergone reaction, and the
mixture was diluted with water (10 mL), cooled in an ice bath,
and acidified with 10% HCl to ca. pH 3. The viscous brown
aqueous solution was extracted with 3:1 CH2Cl2/t-BuOH (5 ×
30 mL) and the combined organic extracts were brine washed
(1 × 25 mL) and dried over MgSO4. Filtration and concentration
yielded a white solid (238 mg, 98%) identified as benzoic acid
via comparison to an authentic sample.
2. NaOH/THF. To a magnetically stirred solution of bromo
ketone 10 (148 mg, 0.800 mmol) in THF (2 mL) at 0 °C was
added powdered NaOH (130 mg) all at once. The mixture was
allowed to warm to ambient temperature for several hours, fitted
with a refux condenser, and refluxed for 4 days. The THF was
evaporated and the residue diluted with 15 mL of water and 15
mL of 3:1 CH2Cl2/t-BuOH. After cooling in a ice bath, the
mixture was acidified to ca. pH 3-4 with 10% HCl. The layers
were separated, and the organic layer was extracted with 3:1
CH2Cl2/t-BuOH (3 × 10 mL). Combined organics were brine
washed (1 × 20 mL), dried over MgSO4, and concentrated to
give 94.1 mg (96%) of a white solid identified as benzoic acid
via comparison to an authentic sample.
Ack n ow led gm en t. This research was supported by
an award from Research Corporation. M.A.F. also
gratefully acknowledges partial financial support by a
Summer Research Grant administered by the Board on
Faculty Research and Development at St. J oseph’s
University.
Su p p or t in g In for m a t ion Ava ila b le: The Gaussian 94
archive entries are available (3 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of this journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
J O951672O