Base-induced solvolyses of [3.2.1]bicyclic α,α′-dichloro ketones - 1,3-transposition and ring-contraction

Article abstract of DOI:10.1002/1099-0690(200111)2001:22<4357::AID-EJOC4357>3.0.CO;2-O

2,4-Dichlorobicyclo[3.2.1]oct-6-en-3-one (1) and its spirocyclopropane (4) and 8-oxa analogues (15, 16) react with methanolic sodium methoxide to furnish bicyclic α-oxo-acetals (3b, 5b, 17, 18) through an enolization/ionization mechanism. With trifluoroethanol/sodium trifluoroethoxide, the corresponding trifluoroethyl acetals (3a and 5a) are formed. Basic hydrolysis affords 2-endo-hydroxynorbornene-2-exo-carboxylic acid (20x) and the 7-oxa analogues (22x, 23x), presumably through benzilic acid rearrangement of the α-diketones.

Full text of DOI:10.1002/1099-0690(200111)2001:22<4357::AID-EJOC4357>3.0.CO;2-O

FULL PAPER
Base-Induced Solvolyses of [3.2.1]Bicyclic α,αЈ-Dichloro Ketones ؊
1,3-Transposition and Ring-Contraction
Baldur Föhlisch,*[a] Andreas Radl,[a][‡] Rüdiger Schwetzler-Raschke,[a][‡] and
Sonja Henkel[a][‡‡]
Keywords: Carbocycles / Ketones / Rearrangements / Ring contractions / Solvolysis
2,4-Dichlorobicyclo[3.2.1]oct-6-en-3-one (1) and its spirocy-
clopropane (4) and 8-oxa analogues (15, 16) react with meth-
anolic sodium methoxide to furnish bicyclic α-oxo-acetals
(3b, 5b, 17, 18) through an enolization/ionization mechanism.
With trifluoroethanol/sodium trifluoroethoxide, the corres-
ponding trifluoroethyl acetals (3a and 5a) are formed. Basic
hydrolysis affords 2-endo-hydroxynorbornene-2-exo-carb-
oxylic acid (20x) and the 7-oxa analogues (22x, 23x), presum-
ably through benzilic acid rearrangement of the α-diketones.
Introduction
It is widely known that the products of the reaction of
α-halogenated ketones with oxygen nucleophiles and bases
depend on many factors and are difficult to predict.[4] This
is in particular the case with dihalo ketones. Concerning the
saturated bicyclo[3.2.1]octane skeleton, it has been reported
that 2-bromobicyclo[3.2.1]octan-3-one, on treatment with
methanolic sodium methoxide, forms the bicyclic methoxy
ketone and methyl norbornane-2-carboxylate, the product
ratio depending on base concentration.[5,6] An enolization-
ionization mechanism involving oxyallyl intermediates was
postulated, and the ring contraction (Favorskii rearrange-
ment) explained by cleavage of a cyclopropanone interme-
diate.[6] To the best of our knowledge, no solvolysis studies
of (oxa)bicyclic α,αЈ-dihalo ketones were available, and so
we investigated the behaviour of 1 under forced conditions
of its formation.
1,1,3,3-Tetrabromoacetone and 1,1,3-trichloroacetone
(TCA) are well known as precursors for the generation of
1,3-dihalo oxyallyl intermediates, which undergo [4 ϩ 3]
cycloadditions with a host of 1,3-diene systems.[1] In par-
ticular, the reaction between tetrabromoacetone and furans
under reductive conditions has found widespread use in or-
ganic synthesis. The standard methodology is the reduction
(debromination) of the resulting 2,4-dibromo-8-oxabicy-
clo[3.2.1]oct-6-en-3-ones, which results in oxabicycles, un-
substituted at positions 2 and 4.[1,2] To the best of our
knowledge, however, non-reductive transformations of the
dihalobicycles have attracted no attention.
Some years ago it was found that TCA reacts in the pres-
ence of several bases, preferably sodium 2,2,2-trifluoro-
ethoxide in 2,2,2-trifluoroethanol (NaTFE/TFE), through
an enolization-ionization mechanism. The dichlorooxyallyl
intermediate can be trapped by cycloaddition with cyclo-
pentadiene or furan(s) to give 2,4-dichlorobicyclo[3.2.1]oct-
6-en-3-one (1) and the 8-oxa analogues.[3] One may well ask
why these bicyclic α,αЈ-dichloro ketones are stable under
these solvolytic reaction conditions and do not undergo a
further reaction.
Results and Discussion
Reaction between Carbocycles and Alcohols/Sodium
Alkoxides
Cyclopentadiene was made to react with TCA in
NaTFE/TFE over 90 minutes at 0 °C Ǟ room temperature.
The dull yellow reaction mixture, containing the cycload-
ducts 1nn, 1nx, and 1xx,[3] was then briefly heated. After
quenching with water and extraction, the ketoacetal 3a was
isolated by kugelrohr distillation in 53% yield.
Unequivocal confirmation of the acetal and ketone con-
stitution was supplied by the carbonyl absorption at 1740
cmϪ1 (IR spectrum), and the large coupling constant of
16.4 Hz extracted from the 1H NMR spectrum, the value of
which is consistent with methylene protons in the position α
to the carbonyl group. The 13C NMR (DEPT technique)
also indicated the new CH2 group [δ ϭ 44.0, C(4)], as well
as the quaternary ‘‘acetal carbon atom’’ at δ ϭ 101.8 [C(2)]
and the carbonyl carbon atom (δ ϭ 201.9).
[‡]
Taken from the doctoral dissertations of A. R. and R. S., Univ.
Stuttgart, 1992.
[‡‡] X-ray analysis.
[a]
Institut für Organische Chemie der Universität Stuttgart,
Pfaffenwaldring 55, 70569 Stuttgart, Germany
Fax: (internat.) ϩ49 (0)711/6854269
In order to investigate this unexpected transposition of
functional groups, endo,endo-2,4-dichlorobicyclo[3.2.1]oct-
E-mail: baldur.foehlisch@po.uni-stuttgart.de
Eur. J. Org. Chem. 2001, 4357Ϫ4365
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/1122Ϫ4357 $ 17.50ϩ.50/0
4357

B. Föhlisch, A. Radl, R. Schwetzler-Raschke, S. Henkel
FULL PAPER
6-en-3-one (1nn) was isolated and allowed to react with Bicyclic α-Diketones
NaTFE/TFE solution. Treatment of 1nn with one equiva-
The rather high yields obtained by alcoholysis of the
lent of NaTFE in TFE, monitored by GLC, resulted in a
new product after a few minutes. After one hour, a further
four peaks were evident in the GLC. Three days later, the
main component was isolated; it could be identified as the
dichlorides 1 and 4 prompted us to prepare the α-diketones
corresponding to the ketoacetals 3 and 5. Indeed, treatment
of the dimethylacetal 3b with hydrochloric acid, after
workup, chromatography, and kugelrohr distillation, gave
the α-diketone 6 in 92% yield as a bright lemon-coloured,
viscous liquid, which solidified at ca. 0 °C. The structure
was verified by spectroscopy (see Experim. Section) and
treatment with o-phenylenediamine to form a quinoxaline
derivative. With 5b, the same procedure resulted in the spiro
diketone 7.
monosubstitution product 2a by its MS, 1H NMR, and 13
C
NMR spectra, but we were unable to obtain the compound
in pure form.
Mechanism of the Alcoholyses
The products of the alcoholyses seem unexpected at first
sight, but are in line with the enolization-ionization mech-
anism postulated for the parent 2-monobromobicy-
clo[3.2.1]octan-3-one case.[6] The difference lies in the fact
that the presence of a second halo substituent results in 1,3-
transposed functional groups in a highly selective trans-
formation, and the Favorskii rearrangement is suppressed.
Deprotonation of one of the chloro-substituted α-carbon
atoms of the dichloro ketone 1 initiates the reaction cascade
(Scheme 1). The resulting enolate 8 can dissociate to form
the ion-pair 9. Association of an alcohol or alkoxide molec-
ule is possible at both termini of the allyl cation, through
which the two alkoxychloroenols 10 and 11 should result.[7]
The ketone 2, isolated on trifluoroethanolysis of the car-
bocycle 1, is consistent with this interpretation, and the exo
configuration at the alkoxy-substituted α-carbon atom
[C(4)] is not unexpected.[6] Obviously, in the ion pair 9 the
exo face is more accessible than the concave endo face and
allows association with alcohol(ate) molecules.
At longer reaction times, or on heating, the alkoxychloro
ketones 2 gave the geminal dialkoxy ketones 3 with 1,3-
transposed functional groups. In view of the fact that we
observed no α,αЈ-dialkoxy ketones, it may be concluded
that the enolization-ionization mechanism again operates:
enolization in the direction of the alkoxy-substituted car-
bon atom again gives an allylic structure (12), which may
form a second ion-pair (13). Obviously, association of the
alcohol at the substituted allylium terminus is preferred, fi-
nally giving the acetals.[7] However, it cannot be ruled out
that the α,α-dialkoxy ketones may result from nucleophilic
substitution of the geminal disubstituted enols 11.
The presence of a trifluoroethoxy group was indicated by
CϪF couplings in the 13C NMR spectrum. The chemical
shifts of 66.4 and δ ϭ 83.9 proved that the chloro and tri-
fluoroethoxy substituents were not at the same position, but
were bound to different carbon atoms: C(2) and C(4). The
endo orientation of the chloro substituent was demon-
1
strated by the magnitude of the vicinal coupling in the H
NMR (J ϭ 3.3 Hz, δ ϭ 4.78) (cf. the starting material 1nn,
δ ϭ 4.67, J ϭ 3.3 Hz). The chemical shift of δ ϭ 3.77 (dd,
J ϭ 3.4 and 1.8 Hz) was consistent with an endo-proton at
C(4), thus demonstrating the exo position of the trifluoro-
ethoxy substituent at C(4).
The reaction between 1 (mixture of endo/exo isomers)
and two (or more) equivalents of NaTFE was complete
after three days at ambient temperature, as indicated by the
precipitation of the stoichiometric amount of NaCl. The
ketoacetal 3a was isolated in a 99% yield. Treatment of 1
with an excess of methanolic sodium methoxide was more
rapid, reaching completion after one hour, and eventually
gave the dimethoxy acetal 3b in the same yield. The spiro-
cyclic ethylene acetal 3c was obtained on treatment of 1
with a solution of sodium in ethylene glycol, using diethyl
ether as cosolvent (yield 72%). Treatment of endo,endo-2,4-
dichlorospiro[bicyclo[3.2.1]oct-6-ene-8,1Ј-cyclopropan]-3-
one (4nn)[3] with NaTFE or NaOMe gave the ketoacetals
5a and 5b, respectively.
Alcoholyses of Oxabicycles
Treatment of dichloro ketone 15, the 8-oxa analogue of
1, with sodium methoxide in methanol at 0 °C Ǟ room
temperature (3 days), followed by extraction and chromato-
graphy, gave the ketoacetal 17. However, the product proved
to be temperature-sensitive, and the yield was mediocre
(38%). Less labile products were isolated on long-term reac-
tion (ca. 1 week) of the saturated oxabicycles 16a and 16b,[8]
which gave the ketoacetals 18a and 18b (21%, 67%). In ana-
logy to 2, exposure of 16b to sodium methoxide solution
4358
Eur. J. Org. Chem. 2001, 4357Ϫ4365

Base-Induced Solvolyses of [3.2.1]Bicyclic α,αЈ-Dichloro Ketones
FULL PAPER
Solvolysis in Water/THF: Ring Contraction
α,αЈ-Dibromocyclohexanones have long been known to
undergo rearrangement on treatment with potash lye, with
formation of 1-hydroxycyclopentanecarboxylic acids.[9] This
ring contraction starts from intermediate 2-hydroxycy-
clohex-2-enones[10] (or rather their 1,2-dioxo tautomers, α-
diketones), which undergo benzilic acid rearrangement,[11]
initiated by the attack of hydroxide anion on the carbonyl
groups. Utaka, Matsushita, and Takeda reported that
cyclohexane-1,2-diones could be obtained in high yield
when THF was used as a cosolvent and the reaction was
conducted with aqueous sodium hydroxide at 0 °C.[12]
With the goal of obtaining these α-diketones directly, we
subjected the bicyclic dichloro ketones 1, 15, 16a, and 16b
to the Utaka conditions. The resulting products were non-
volatile, giving no GLC signals. On acidification with hy-
drochloric acid, no (tautomeric) α-diketones were isolated,
but (hydroxy)carboxylic acids were found. In other words,
ring contraction had occurred even under these mild condi-
tions. In the case of 1, both the endo and the exo isomers
(1nn and 1xx) were found to give practically the same re-
sult. Examination of the spectra (see Experim. Section) in-
dicated that one of the diastereomeric 2-hydroxynorbor-
nene-2-carboxylic acids (20), with a m.p. of 108Ϫ109 °C,
must have been formed from 1.
Scheme 1. Proposed alcoholysis mechanism with the dichloro
ketone 1
for a more usual, shorter reaction time (100 min) resulted
in the exchange of only one chloro atom, producing the
chloro-substituted monomethoxy ketone 19a in 55% yield.
Alder, Hartmann, and Roth reported that the exo-carb-
oxylic acid (20x) was formed on heating 2endo-bromonor-
bornene-2exo-carboxylic acid (21x) with 10 % aqueous so-
dium carbonate, in addition to small amounts of a hydroxy-
nortricyclenecarboxylic acid.[13] However, these authors re-
ported a melting point of 145 °C for 20x. Spectra were not
mentioned; NMR, of course, was not available in those
days. To the best of our knowledge, the endo-carboxylic acid
(20n) was not known in the literature. The different melting
points Ϫ 108 °C versus 145 °C Ϫ led us for the moment to
the conclusion that 20n was the formula of the compound
formed from 1: namely the endo-carboxylic acid. We there-
fore expected that the compound would undergo iodolac-
tonization.[14] However, to our surprise, we could not suc-
ceed in obtaining an iodolactone on treatment of our acid
20 in aqueous sodium carbonate solution with the iodine/
potassium iodide reagent.[15]
Eur. J. Org. Chem. 2001, 4357Ϫ4365
4359

B. Föhlisch, A. Radl, R. Schwetzler-Raschke, S. Henkel
FULL PAPER
Figure 1. X-ray structure of the 2-hydroxynorbornene-2-carboxylic acid with m.p. 108Ϫ109 °C (20x): stereoscopic projection of the single
molecule; see also refs.[16,17]
In order to identify the acids by spectral comparison, we 23ax had the same configuration at C(2), the unsaturated
tried to prepare 20x or 20n by the procedure given by Alder acid was hydrogenated over palladium-charcoal catalyst.
et al.[14] However, we were unable to confirm Alder’s re- The products were found to be identical according to their
sults.[16] Finally, the structure was determined by X-ray ana- 1H NMR spectra.
lysis, which showed the exo position of the COOH group
When 16b was allowed to hydrolyse at 0 °C (ice bath) for
4 hours, the monochloro hydroxy ketone 19b Ϫ an analogue
and the endo position of the hydroxy group (Figure 1).[17]
The 8-oxa analogue 15, on treatment with aqueous of the oxabicycle 19a (OH instead of OCH3) Ϫ could be
NaOH/THF, gave an unsaturated oxanorbornenecarboxylic isolated in 37% yield. This demonstrates that at least the
acid (22) in a disappointing 19% yield: again, only one of first step of the hydrolysis and the aforementioned alco-
the possible stereoisomers. Since we were once more unable holysis reactions was the same.
to obtain an iodolactone, formula 22x seems likely. From
Mechanism
the 1,5-dimethyl analogue, endo,endo-2,4-dichloro-1,5-di-
methyl-8-oxabicyclo[3.2.1]oct-6-en-3-one,[3] no acidic prod-
uct could be extracted from the reaction mixture.
The different behaviour of the dichloro ketones 1, 15, and
16 on basic hydrolysis and alcoholysis is striking at first
Better results were obtained with the saturated oxa- sight. Whereas in the latter case the carbon skeleton is re-
bicycles 16a and 16b. The oxanorbornanecarboxylic acids tained, ring contraction occurs in the aqueous solvent. The
23ax and 23bx were isolated in 37% and 60% yields, re- discrepancy was resolved by our finding that the diketone
spectively. In order to demonstrate that the acids 22x and 6 rearranged to 20x in alkaline aqueous media.
Scheme 2. Benzilic acid rearrangement of diketone 6
4360
Eur. J. Org. Chem. 2001, 4357Ϫ4365

Base-Induced Solvolyses of [3.2.1]Bicyclic α,αЈ-Dichloro Ketones
FULL PAPER
or
Polygram
N/UV254
(alumina),
distributed
by
The outcome is explained by attack of hydroxide anion
at C(3) (Scheme 2). However, exo-attack of hydroxide
would be expected to produce an intermediate with a chair
conformation; assuming the anionotropic 1,2-shift to be
concerted, one would anticipate formation of the endo-
carboxylic acid 20n. The formation of 20x may be explained
by a boat conformer intermediate arising from ring-inver-
sion or endo-attack of hydroxide.[18]
MachereyϪNagel & Co, Düren, Germany; detection by UV extinc-
tion, exposure to iodine vapour, or by spraying with vanillin/H2SO4
solution, followed by warming. Ϫ Preparative column chromato-
graphy: silica gel 60 (63Ϫ200 µm), distributed by
MachereyϪNagel & Co, Düren, Germany. Alumina B was supplied
by ICN Biomedicals, Eschwege, Germany. Ϫ Dry petroleum ether
(PE) was distilled (b.p. 40Ϫ65 °C). Ethyl acetate (EA) was dried
˚
over calcium chloride, distilled, and kept dry over 4-A molecular
It is therefore probable that on basic hydrolysis the
sieves. Ϫ Melting points were determined with a Büchi 510 appar-
dichloro ketones 1, 15, and 16 at first give the diketones atus (Büchi Laboratoriumstechnik AG, Flawil/Switzerland) and are
not corrected. Ϫ Elemental analyses were performed by the service
of the Institut für Organische Chemie, University of Stuttgart. Ϫ
Dichloromethane was dried by refluxing over powdered calcium
hydride and distilled. For sodium trifluoroethoxide/trifluoroe-
thanol (NaTFE/TFE) reagent see ref.[8] Palladium on activated
charcoal (10%) hydrogenation catalyst was supplied by Janssen, Be-
erse, Belgium. 1,1,3-Trichloroacetone (TCA, 1,1,3-trichloropropan-
2-one) was obtained from Wacker Chemie, Munich, Germany in
‘‘technical’’ grade. It was distilled prior to use over a 25 cm Vigreux
column in vacuo (b.p. 89Ϫ92/75 Torr).
such as 6, following the mechanism shown in Scheme 1
(with H2O instead of ROH) and in analogy with the mono-
cyclic system: α,αЈ-dibromocyclohexanone. The difference
may be that cyclohexane-1,2-dione would be stabilized as
the monoanion by rapid enolization in alkaline water; enol-
ization of the bicyclic diketones may be impeded by ring
strain, opening a channel for nucleophilic attack on the car-
bonyl groups by hydroxide ion.[19] As a matter of course, a
Favorskii rearrangement of the dichloro ketones 1, 15, and
16 by way of the cyclopropanone mechanism[6] cannot be
rigorously ruled out. A further possibility, the semibenzilic
acid pathway, is not very probable since 1 Ϫ according to
Scheme 1 Ϫ reacts rapidly through the enolate, and the se-
mibenzilic mechanism is known to be followed in cases in
which the α-halo ketone lacks enolizable αЈ-carbon
Treatment of Dichlorobicyclo[3.2.1]oct-6-en-3-one (1nn) with
NaTFE (1 equivalent).
؊ 2-endo-Chloro-4-exo-(2,2,2-trifluoro-
ethoxy)bicyclo[3.2.1]oct-6-en-3-one (2a): A solution of NaTFE in
TFE (c ϭ 2.0 mol/L, 5 mL) was added dropwise with magnetic
stirring to a solution of 1nn[3] (1.91 g, 10.0 mmol) in TFE (10 mL).
The mixture was stirred for 3 days at room temperature. Water
atoms.[20,21] Moreover, in this case, it would be anticipated (30 mL) was added, and the mixture was extracted with dichloro-
methane (6 ϫ 10 mL). The combined extracts were dried with mag-
nesium sulfate and concentrated in a rotary evaporator. The yellow,
oily residue (2.83 g) was distilled in a kugelrohr apparatus at 70Ϫ80
°C/0.005 Torr. The colourless liquid (2.15 g) consisted mainly of 2a,
together with 3a and further unidentified components (ca. 10% by
GLC). The yield of 2a was ca. 76%. The following data for 2a
were extracted from the spectra of the mixture: 13C NMR/DEPT
(62.9 MHz, CDCl3): δ ϭ 38.35, 44.2, 47.7, 66.4, 67.4 (q, J ϭ
that 2-chloronorbornene-1-carboxylic acid would be
formed from 1, followed by hydrolysis to 2-hydroxynor-
bornene-1-carboxylic acid and/or a hydroxynortricyclene-
carboxylic acid,[13,16] rather than 20x.
Conclusion
1
35.0 Hz), 83.9, 123.35 (q, J ϭ 278 Hz), 133.3, 137.9, 198.7. Ϫ H
Apart from providing a deeper mechanistic insight into
the behaviour of (bi)cyclic α,αЈ-dihalo ketones under solvo-
lytic conditions, the ketoacetals and α-diketones are anticip-
ated to be useful as building blocks in organic synthesis,
and so should stimulate further research.[22] Presumably,
optimization of the conditions of alcoholysis and workup
for the oxabicycles 16؊18 would give better yields. It may
also be anticipated that the corresponding α,αЈ-dibromo ke-
tones, available by cyclocondensation with tetrabromoace-
tone under reductive conditions (see, for example, ref.[2]),
would give the same results. The ring contraction to the
norbornene system should be exploitable too, especially
with the spiro compound(s).[23]
NMR (250 MHz, CDCl3): δ ϭ 2.11 (mc, 1 H), 2.46 (d, J ϭ 12.1 Hz,
1 H), 3.11 (mc, 1 H), 3.17 (mc, 1 H), 3.77 (dd, J ϭ 3.4, J ϭ 1.8, 1
H), 3.91 (m, 2 H), 4.78 (d, J ϭ 3.3 Hz, 1 H), 6.06 (dd, J ϭ 5.8,
˜
J ϭ 2.9, 1 H), 6.35 (dd, J ϭ 5.8, J ϭ 2.1, 1 H). Ϫ IR (film): ν ϭ
1735 cmϪ1 (CϭO). Ϫ EIMS (70 eV): m/z (%) ϭ 318 (5) [Mϩ from
C12H12F6O3 (3a)], 256 (6) [Mϩ from C10H1037ClF3O2], 254 (19)
[Mϩ from C10H1035ClF3O2], 66 (100) [C5H6ϩ]. Ϫ HRMS: calcd.
for C10H1035ClF3O2 254.0321; found 254.0323.
2,2-Bis(2,2,2-trifluoroethoxy)bicyclo[3.2.1]oct-6-en-3-one
(3a):
a) Cyclopentadiene (1.32 g, 20.0 mmol), freshly prepared by crack-
ing distillation of dicyclopentadiene, was mixed with 1,1,3-trichloro-
propan-2-one (TCA) (1.61 g, 10 mmol) and cooled in an ice bath.
A solution of NaTFE in TFE (c ϭ 2.0 mol/L, 20 mL) was added
dropwise at 0 °C with magnetic stirring. A white precipitate (NaCl)
appeared. The mixture was stirred at room temperature for one
hour. The dull yellow mixture was then refluxed for 5Ϫ10 minutes,
whereupon the colour turned brown. When the mixture had cooled
to room temperature, water (50 mL) was added and the mixture
was extracted with dichloromethane (5 ϫ 25 mL). The combined
organic layers were dried with magnesium sulfate and concentrated
in a rotary evaporator. The dark, oily residue was distilled in a
kugelrohr apparatus at 90Ϫ100 °C/0.03 Torr, giving 1.67 g of a pale
yellow oil; yield 53%.
Experimental Section
General Remarks: IR: PerkinϪElmer 457. Ϫ NMR: Bruker AC 250
1
for 62.9 MHz 13C NMR and 250 MHz H NMR spectra, Bruker
1
CXP 300 for 300 MHz H NMR and 75.47 MHz 13C NMR spec-
tra. CDCl3 or [D6]DMSO as solvent, TMS as internal standard.
Assignment of the 13C NMR signals of the cycloadducts and their
transformation products was made with off-resonance or DEPT
spectra. Ϫ EIMS: Varian MAT 711 with data system SS 100. Ϫ
b) A solution of NaTFE in TFE (c ϭ 2.0 mol/L, 6 mL) was added
Analytical TLC: precoated sheets, Polygram Sil G/UV254 (silica), to a solution of 1 [mixture of isomers, 0.96 g, 5 mmol)] in TFE
Eur. J. Org. Chem. 2001, 4357Ϫ4365 4361

B. Föhlisch, A. Radl, R. Schwetzler-Raschke, S. Henkel
FULL PAPER
(10 mL), and the mixture was stirred at room temperature for 3
CDCl3): δ ϭ 0.52Ϫ0.79 (m, 2 H), 1.07Ϫ1.15 (m, 2 H), 2.12 (dd,
days. Water (30 mL) was added, and the mixture was extracted with J ϭ 5.8, J ϭ 2.9, 1 H), 2.46 (d, J ϭ 3.0 Hz, 1 H), 2.50 (dd, J ϭ
dichloromethane (6 ϫ 10 mL). The combined extracts were dried 16.0, J ϭ 3.8, 1 H), 2.68 (dd, J ϭ 16.9, J ϭ 3.5, 1 H), 3.70Ϫ4.04
with magnesium sulfate and concentrated in a rotary evaporator.
Kugelrohr distillation of the remaining yellow oil (1.81 g) at 70Ϫ80
(m, 4 H), 6.02 (dd, J ϭ 5.9, J ϭ 2.9, 1 H), 6.31 (dd, J ϭ 5.9, J ϭ
2.8, 1 H). Ϫ 13C NMR/DEPT (62.9 MHz, CDCl3): δ ϭ 7.7, 11.4,
°C/0.001 Torr gave 1.58 g (99%) of 3a, as a yellow oil. Ϫ IR (film): 32.7, 43.5, 45.3, 51.3, 59.6, (q, J ϭ 36.0 Hz), 61.0 (q, J ϭ 35.0 Hz),
1
ν˜ ϭ 1740 cmϪ1. Ϫ H NMR (250 MHz, CDCl3): δ ϭ 2.15 (mc, 2
100.6, 123.6 (q, J ϭ 276 Hz), 130.55, 140.85, 201.3. A satisfactory
H), 2.40 (mc, 1 H), 2.64 (dd, J ϭ 16.4, J ϭ 3.2, 1 H), 2.91 (mc, 1 combustion analysis could not be obtained.
H), 2.91 (mc, 1 H), 3.07 (mc, 1 H), 3.92 (m, 2 H), 4.31 (dq, J ϭ
2,2-Dimethoxyspiro[bicyclo[3.2.1]oct-6-ene-8,1Ј-cyclopropan]-3-one
12.1, J ϭ 8.6, 2 H), 5.90 (dd, J ϭ 5.7, J ϭ 3.0, 1 H), 6.24 (dd, J ϭ
5.7, J ϭ 2.8, 1 H). Ϫ 13C NMR/DEPT (62.8 MHz, CDCl3): δ ϭ
38.3, 44.0, 45.9, 60.1 (q, J ϭ 35.0 Hz), 61.1 (q, J ϭ 35.0 Hz),
101.75, 123.5, 123.7 (q, J ϭ 277 Hz), 130.1, 140.2, 201.9. Ϫ
C12H12F6O3 (318.2) calcd. C 45.30, H 3.80; found C 45.26, H 3.73.
(5b) A solution of sodium methoxide in methanol (c ϭ 2.0 mol/L,
25 mL) was added with stirring over 15 min to a solution of 4nn
(2.17 g, 10.0 mmol) in dry methanol (50 mL), cooled in an ice bath.
The cooling bath was removed, and the mixture was stirred for 1
hour. Water was added, and the mixture was extracted with diethyl
ether (5 ϫ 50 mL). The combined ether extracts were dried with
2,2-Dimethoxybicyclo[3.2.1]oct-6-en-3-one (3b): A solution of so-
dium methoxide in methanol (c ϭ 2.0 mol/L, 24 mL) was added magnesium sulfate and concentrated in a rotary evaporator. Kugel-
with stirring over 5 min to a cooled solution (Ϫ14 °C) of 1 (mixture
rohr distillation of the residue at 75Ϫ85 °C/0.001 Torr gave 1.85 g
˜
(89%) of 5b, as a pale yellow viscous oil. Ϫ IR (film): ν ϭ 1725
of isomers, 1.91 g, 10.0 mmol) in dry methanol (50 mL). The cool-
1
ing bath was removed, and the mixture was stirred for 1 hour. On cmϪ1 (CϭO). Ϫ H NMR (250 MHz, CDCl3): δ ϭ 0.67Ϫ0.75 (m,
warming to room temperature, a white precipitate (NaCl) formed. 2 H), 1.08Ϫ1.18 (m, 2 H), 2.07 (mc, 1 H), 2.41 (dd, J ϭ 16.1, J ϭ
Water was added, and the mixture was extracted with diethyl ether
2.9, 1 H), 2.50 (mc, 1 H), 2.64 (dd, J ϭ 16.0, J ϭ 3.2, 1 H), 3.17
(5 ϫ 50 mL). The combined ether extracts were dried with magnes- (s, 3 H), 3.31 (s, 3 H), 6.03 (dd, J ϭ 5.9, J ϭ 2.9, 1 H), 6.28 (dd,
ium sulfate and concentrated in a rotary evaporator. Kugelrohr dis-
tillation of the remaining yellow oil (1.93 g) at 80Ϫ90 °C/0.005 Torr
J ϭ 5.9, J ϭ 2.8, 1 H). Ϫ 13C NMR/DEPT (62.9 MHz, CDCl3):
δ ϭ 8.3, 11.4, 33.0, 43.6, 45.8, 48.2, 50.8, 50.9, 101.5, 132.1, 139.3,
gave 1.80 g (99%) of 3b as a colourless oil. Ϫ IR (film): ν˜ ϭ 1730 203.8. Ϫ C12H16O3 (208.3): calcd. C 69.21, H 7.74; found C 69.04,
cmϪ1 (CϭO). Ϫ 1H NMR (300 MHz, CDCl3): δ ϭ 2.08 (mc, 1 H), H 7.84.
2.22 (d, J ϭ 11.2 Hz, 1 H), 2.34 (dt, J ϭ 16.0, J ϭ 3.0, 1 H), 2.56
Bicyclo[3.2.1]oct-6-ene-2,3-dione (6): A mixture of 3b (1.82 g,
(dd, J ϭ 16.0, J ϭ 3.1, 1 H), 2.84 (mc, 1 H), 3.16 (dd, J ϭ 5.2, J ϭ
10.0 mmol), water (10 mL), and conc. hydrochloric acid (10 mL)
2.9, 1 H), 3.33 (s, 3 H), 3.38 (s, 3 H), 5.89 (mc, 1 H), 6.17 (mc, 1
was stirred overnight at room temperature. The mixture was ex-
tracted with dichloromethane (6 ϫ 50 mL). The combined extracts
were dried with magnesium sulfate and concentrated in a rotary
H). Ϫ 13C NMR/off-resonance (75.5 MHz, CDCl3): δ ϭ 38.5 (d),
38.75 (t), 44.0 (t), 45.2 (d), 48.8 (q), 50.15 (q), 101.9 (s), 139.0 (d),
203.1 (s). Ϫ C10H14O3 (182.2) calcd. C 65.92, H 7.74; found C
evaporator. The remaining yellow oil (1.60 g) was purified by filtra-
65.96, H 7.62.
tion over silica (80 g) with PE/EA (1:1, 250 mL). The filtrate was
2,2-Ethylenedioxybicyclo[3.2.1]oct-6-en-3-one (3c): A solution of 1
concentrated, and the viscous residue (1.43 g) was distilled in a ku-
(mixture of isomers, 1.91 g, 10.0 mmol) in diethyl ether (10 mL) gelrohr apparatus at 100 °C/0.001 Torr. Yield 1.25 g (92%) 6. The
˜
lemon-coloured, viscous oil solidified at ca. 0 °C. Ϫ IR (film): ν ϭ
was cooled to 0 °C under nitrogen. A solution of sodium ethane-
1,2-diolate in ethane-1,2-diol (c ϭ 2.0 mol/L, 20 mL) was added 1725 cmϪ1 (broad absorption band). Ϫ 1H NMR (300 MHz,
over 15 min, with continuous stirring and cooling. The mixture CDCl3): δ ϭ 2.33 (d, J ϭ 12.1 Hz, 1 H), 2.55Ϫ2.75 (m, 3 H), 3.13
was stirred overnight at room temperature, whereupon the colour
changed to brown. Brine (20 mL) was added, and the mixture was
extracted with dichloromethane (5 ϫ 20 mL). The combined or-
ganic layers were dried with magnesium sulfate. Evaporation of the
solvent and kugelrohr distillation of the remaining brown liquid
(mc, 1 H), 3.58 (dd, J ϭ 4.6, J ϭ 3.1, 1 H), 5.88 (dd, J ϭ 5.4, J ϭ
2.9, 1 H), 6.56 (dd, J ϭ 5.5, J ϭ 2.8, 1 H). Ϫ 13C NMR/off-reson-
ance (75.47 MHz, CDCl3,): δ ϭ 37.7 (d), 39.2 (t), 45.6 (t), 54.7 (d),
127.6 (d), 142.9 (d), 189.5, 197.1. Ϫ EIMS (70 eV): m/z (%) ϭ 136
(12) [Mϩ from C8H8O2], 108 (16) [Mϩ Ϫ CO], 79 (15), 66 (100),
(2.86 g) at 70Ϫ90 °C/0.005 Torr gave 1.29 g (72%) of a yellow oil 39 (11). Ϫ C8H8O2 (136.15): calcd. C 70.58, H 5.92; found C 68.54,
(3c). Ϫ IR (CCl4): ν ϭ 1730 cmϪ1. Ϫ 1H NMR (250 MHz, CDCl3): H 6.10; a satisfactory combustion analysis could not be obtained.
˜
δ ϭ 2.16 (mc, 1 H), 2.24 (mc, 1 H), 2.30 (dt, J ϭ 16.8, J ϭ 2.9, 1 Ϫ Quinoxaline derivative:[16,24] C14H12N2 (208.3): calcd. C 80.74,
H), 2.61 (dd, J ϭ 16.6, J ϭ 3.3, 1 H), 2.79 (mc, 1 H), 2.86 (mc, 1
H), 3.80Ϫ4.40 (m, 4 H), 5.97 (dd, J ϭ 5.7, J ϭ 2.9, 1 H), 6.21 (dd,
J ϭ 5.7, J ϭ 2.7, 1 H). Ϫ 13C NMR/DEPT (62.9 MHz, CDCl3):
δ ϭ 38.2, 39.6, 43.7, 48.3, 64.7, 67.1, 108.1, 131.1, 139.7, 206.6. Ϫ
C10H12O3 (180.2) calcd. C 66.65, H 6.71; found C 66.94, H 6.83.
H 5.81, N 13.45; found C 80.50, H 5.85, N 13.45.
Spiro[bicyclo[3.2.1]oct-6-ene-8,1Ј-cyclopropane]-2,3-dione (7): Com-
pound 5b (1.46 g, 7 mmol) was stirred with water (10 mL) and
conc. hydrochloric acid (10 mL) as described for 6. After 2 days,
the mixture was diluted with water (10 mL) and extracted with
2,2-Bis(2,2,2-trifluoroethoxy)spiro[bicyclo[3.2.1]oct-6-ene-8,1Ј- dichloromethane (5 ϫ 50 mL). The combined extracts were dried
cyclopropan]-3-one (5a): A solution of NaTFE in TFE (c ϭ 2.0
with magnesium sulfate and concentrated in a rotary evaporator.
mol/L, 6 mL) was added dropwise to a solution of 4nn[3] (2.17 g, The residue (1.30 g) was purified by filtration over silica (80 g) with
10.0 mmol) in TFE (10 mL), and the mixture was stirred at room PE/EA (4:1). The filtrate was concentrated, and the remaining yel-
temperature for 3 days. Brine (20 mL) was added, and the precipit-
low solid (1.15 g) was sublimed at 50Ϫ60 °C/0.001 Torr. The yellow
ate was dissolved by adding water. The mixture was extracted with sublimate (1.12 g, yield 99%) 7 had m.p. 82Ϫ84 °C. Ϫ IR (CDCl3):
dichloromethane (5 ϫ 25 mL). The combined extracts were dried
with magnesium sulfate and concentrated in a rotary evaporator.
ν˜ ϭ 1740, 1720 cmϪ1. Ϫ 1H NMR (250 MHz, CDCl3): δ ϭ 0.75
(mc, 4 H), 2.44 (mc, 1 H), 2.64 (dd, J ϭ 18.3, J ϭ 2.0, 1 H), 2.75
Kugelrohr distillation of the remaining yellow oil (3.40 g) at 70Ϫ90 (dd, J ϭ 18.3, J ϭ 4.1, 1 H), 2.97 (d, J ϭ 3.0 Hz, 1 H), 6.00 (dd,
°C/0.001 Torr gave 2.89 g (84%) of 5a, as a yellow viscous oil that
J ϭ 5.7, J ϭ 3.0, 1 H), 6.62 (dd, J ϭ 5.7, J ϭ 2.9, 1 H). Ϫ 13C
solidified on storage in the refrigerator. Ϫ 1H NMR (250 MHz, NMR/DEPT (62.9 MHz, CDCl3): δ ϭ 5.4, 11.5, 34.0, 44.65, 45.4,
4362
Eur. J. Org. Chem. 2001, 4357Ϫ4365

Base-Induced Solvolyses of [3.2.1]Bicyclic α,αЈ-Dichloro Ketones
61.8, 127.9, 142.9, 188.4, 198.3. Ϫ EIMS (70 eV); (m/z): 162 (6)
FULL PAPER
the lines are split into doublets with J ϭ 1.4 Hz, 1 H, 4x-H), 3.31
[Mϩ from C10H10O2], 134 (5) [M Ϫ CO], 106 (4) [M Ϫ 2 CO], 93 (s, 3 H, CH3O), 3.47 (s, 3 H, CH3O). Ϫ 13C NMR (75.47 MHz,
(8), 92 (100) [M Ϫ C3H2O2]. Ϫ C10H10O2 (161.2): calcd. C 74.06,
CDCl3): δ ϭ 20.65, 25.8, 33.1, 36.4, 51.6 (CH3O), 52.3 (CH3O),
H 6.21; found C 73.25, H 6.22. A satisfactory combustion analysis 54.0, 82.05, 86.45, 101.04, 204.2. Ϫ C11H18O4 (214.3): calcd. 61.66,
could not be obtained. Ϫ HRMS: calcd. for C10H10O2 162.0681;
H 8.47; found C 61.45, H 8.31.
found 162.0682.
2-Chloro-4-methoxy-1,5-dimethyl-8-oxabicyclo[3.2.1]octan-3-one
(19a): This compound was prepared from 18b (2.23 g, 10.0 mmol)
in methanol (60 mL) and sodium methoxide solution (50 mL, c ϭ
1 mol/L) at 0 °C (80 min). After stirring at room temperature for 20
min and workup as usual, the residue was distilled in a kugelrohr
apparatus at 120 °C/0.01 Torr, giving 1.73 g of a yellow oil. It was
purified by filtration over alumina (activity 3) with PE/EA (10:1).
The eluate, a colourless oil, was concentrated and distilled in a
kugelrohr apparatus at 130 °C/0.02 Torr. Yield 1.20 g (55%), as a
2,2-Dimethoxy-8-oxabicyclo[3.2.1]oct-6-en-3-one (17): A solution of
15 (1.93 g, 10.0 mmol) in dry methanol (50 mL) was added drop-
wise over 15 minutes to a solution of sodium methoxide in dry
methanol (c ϭ 1 mol/L, 50 mL), cooled in an ice bath. The ice
bath was removed, and the mixture was stirred for 3 days at room
temperature. After addition of water (60 mL), the mixture was con-
centrated by half in a rotary evaporator and extracted with ethyl
acetate (7 ϫ 20 mL). The combined extracts were dried with mag-
nesium sulfate, and the solvent was removed by rotary evaporation.
The residue (1.31 g), a brown oil, was purified by chromatography
on alumina (activity 3, 200 g) with PE/EA (5:1). The eluted product
with Rf ϭ 0.21 was dried in vacuo (oil pump). The substance, a
pale yellow oil (0.70 g 17, 38%) proved to be very temperature-
sensitive and so was not subjected to kugelrohr distillation. Ϫ IR
colourless oil. Ϫ IR (film): ν˜ ϭ 1735 with shoulder at 1725 cmϪ1
.
Ϫ
1H NMR (300 MHz, CDCl3): δ ϭ 1.37Ϫ2.08 (m, 10 H, sur-
mounted by two singlets at δ ϭ 1.46 and 1.57 (CH3), 3.32 (s, 1 H,
4-H), 3.39 (s, 3 H, CH3O), 4.67 (d, J ϭ 1.4 Hz, 1 H, 2x-H). Ϫ 13C
NMR/off-resonance (62.8 MHz, CDCl3): δ ϭ 20.8 (q), 24.15 (q),
31.0 (t), 33.4 (t), 58.4 (q, CH3O), 68.8 (d), 83.6 (s), 86.95 (s), 89.7
(d), 199.0 (s, C-3). Ϫ EIMS (20 eV): m/z (%) ϭ 218 (5) [Mϩ], 184
(10), 183 (98), 155 (11), 154 (76), 151 (16), 142 (14), 139 (17), 122
(12), 118 (18), 109 (48), 99 (11), 98 (16), 97 (100), 96 (23), 72 (61),
(film): ν ϭ 1735 cmϪ1. Ϫ 1H NMR (80 MHz, CDCl3): δ ϭ 2.31
˜
(dd, J ϭ 15.0, J ϭ 1.0, 1 H, 4n-H), 2.95 (dd, J ϭ 15.0, J ϭ 4.5, 1
H, 4x-H), 3.36 (s, 3 H, CH3O-), 3.50 (s, 3 H, CH3O-), 4.98Ϫ5.04
(m, 2 H, 1-H and 5-H), AB sub-spectrum centred at δ ϭ 6.30, with
δA ϭ 6.37 and δB ϭ 6.22, J ϭ 6.0 Hz; the lines of the A- and B-
parts are split into doublets with J ϭ 1.5 Hz, 6-H and 7-H). Ϫ 13C
NMR/off-resonance (75.47 MHz, CDCl3): δ ϭ 44.8 (t, C-4), 50.1
(q, CH3O), 50.9 (q, CH3O), 78.3 (d, C-5), 80.2 (d, C-1), 100.6 (s,
C-2), 129.8 (d, C-6 or C-7), 136.0 (d, C-6 or C-7), 200.8 (s, C-3).
Ϫ C9H12O4 (184.2): calcd. C 58.69, H 6.57; found C 58.55, H 6.49.
43 (50).
218.0708.
Ϫ HRMS. calcd. for C10H15ClO3 218.0710; found
2-endo-Hydroxybicyclo[2.2.1]hept-5-ene-2exo-carboxylic Acid (20x):
a) Compound 1nn (0.96 g, 5.00 mmol) was dissolved in a stirred
mixture of THF (5 mL) and water (10 mL). A solution of sodium
hydroxide (3.00 g, 75.0 mmol) in water (15 mL) was added drop-
wise under nitrogen over 30 min and with cooling in an ice bath.
The colour of the mixture turned pale yellow. Stirring was con-
tinued for 1 h at 0 °C and 3 h at room temperature. The mixture
was acidified by addition of a mixture of conc. hydrochloric acid
(10 mL) and water (25 mL) and stirred overnight. It was extracted
with ethyl acetate (10 ϫ 20 mL). The combined organic layers were
dried with magnesium sulfate, and the solvent was removed in a
rotary evaporator. The resinous residue (0.97 g) was dissolved in
dichloromethane (100 mL). Slow evaporation at room temperature
in a flask, covered with a watch glass, gave (after 4 weeks) 0.67 g
of colourless crystals with m.p. 108Ϫ109 °C (87% yield). From the
first, square-shaped crystals (appearing after one week) one (with
a length of ca. 2 mm) was selected and cut into two halves with a
razor blade. One half was taken for the X-ray crystal structure ana-
lysis.[17]
2,2-Dimethoxy-8-oxabicyclo[3.2.1]octan-3-one (18a): This com-
pound was prepared from 16a (1.95 g, 10.0 mmol) as described for
17 (see above). Treatment was at 0 °C for 1 h, then 7 days stirring
at room temperature. The same workup and chromatography pro-
vided a colourless oil (Rf ϭ 0.28) that crystallized in the refriger-
˜
ator. Yield 0.39 g (21%) with m.p. 42Ϫ43 °C. Ϫ IR (KBr): ν ϭ
1
1735 cmϪ1. Ϫ H NMR (300 MHz, CDCl3): δ ϭ 1.62Ϫ2.08 (m, 4
H, 6-H and 7-H), 2.30 (finely split d, J ϭ 14.6 Hz, 1 H, 4n-H),
2.93 (dd, the low-field part of the doublets is finely split, J ϭ 14.6,
J ϭ 4.9, 1 H, 4x-H), 3.26 (s, 3 H, CH3O), 3.38 (s, 3 H, CH3O),
4.60Ϫ4.62 (m, 1 H, 1-H or 5-H), 4.68Ϫ4.71 (m, 1 H, 1-H or 5-H).
Ϫ
13C NMR (75.47 MHz, CDCl3): δ ϭ 23.5 (C-7), 28.6 (C-6), 48.2
(C-4), 49.5 and 50.3 (CH3O), 75.8 (C-5), 77.9 (C-1), 100.3 (C-2),
202.9 (C-3). Ϫ C9H14O4 (186.2): calcd. C 58.05, H 7.58; found C
58.20, H 7.57.
b) The same procedure was performed with 0.96 g of 1xx. Yield
0.64 g (83%) of colourless crystals with m.p. 108 °C. The spectra
were identical with those from the substance obtained by proced-
ure a).
2,2-Dimethoxy-1,5-dimethyl-8-oxabicyclo[3.2.1]octan-3-one (18b): A
solution of 16b (2.23 g, 10.0 mmol) in dry methanol (80 mL) was
added dropwise over 20 minutes to a solution of sodium methoxide
in dry methanol (c ϭ 1 mol/L, 45 mL), cooled in an ice bath. The
mixture was stirred for 1 h in the ice bath and then for 5 days at
c) A solution of potassium hydroxide (2.00 g) in water (10 mL) was
room temperature. After addition of water (30 mL), the mixture added to diketone 6 (1.36 g, 10.0 mmol) . The yellow mixture dis-
was concentrated by half in a rotary evaporator, and extracted with coloured and finally took on a brownish colour. The mixture was
dichloromethane (5 ϫ 25 mL). The combined extracts were dried
acidified with 10% hydrochloric acid (10 mL), stirred overnight,
with magnesium sulfate, and the solvent was removed by rotary and worked up as described above; the resinous residue (1.51 g)
evaporation. The residue was distilled in a kugelrohr apparatus at
150 °C/0.01 Torr, to give 1.44 g (67%) of 18b, as a colourless oil.
For analysis, a sample was purified by chromatography on alumina
was dissolved in ethyl acetate (10 mL). Petroleum ether (20 mL)
was added dropwise, and the mixture was stored in the refrigerator.
Colourless crystals with m.p. 108 °C were deposited (1.03 g, 67%
(activity 3) with PE/EA (10:1). The eluted product with Rf ϭ 0.25 yield). The NMR spectra were in agreement with those of the sub-
1
was dried in vacuo (oil pump). Ϫ IR (KBr): ν ϭ 1735 cmϪ1. Ϫ H stances described above. Ϫ IR (KBr): ν ϭ 3440 (OH), broad ab-
˜
˜
NMR (300 MHz, CDCl3): δ ϭ 1.42 (s, 3 H, CH3), 1.47 (s, 3 H, sorption band at ca. 3000 with peaks at 3070 (ϭCH), 2990, 2940;
1
CH3), 1.61Ϫ1.83 (m, 3 H, 6-H and 7-H), 2.00Ϫ2.11 (m, 1 H, 6-H 1700 cmϪ1 (broad, COOH). Ϫ H NMR (300 MHz, CDCl3): δ ϭ
or 7x-H), 2.38 (d, J ϭ 14.3 Hz, 1 H, 4n-H), 2.70 (d, J ϭ 14.3 Hz, 1.26 (dd, J ϭ 12.5, J ϭ 3.7, 1 H), 1.56 (dt, J ϭ 9.1, J ϭ 1.8, 1 H),
Eur. J. Org. Chem. 2001, 4357Ϫ4365
4363

B. Föhlisch, A. Radl, R. Schwetzler-Raschke, S. Henkel
FULL PAPER
2.06 (d, J ϭ 9.1 Hz, 1 H), 2.50 (dd, J ϭ 12.5, J ϭ 3.7, 1 H), 3.02 of hydrogen gas until the uptake of hydrogen stopped (50 min).
(br. s, 1 H), 3.09 (br. s, 1 H), 6.21 (dd, J ϭ 5.6, J ϭ 3.0, 1 H), 6.55 The mixture was filtered, and the catalyst was washed with ethyl
(dd, J ϭ 5.6 Hz, J ϭ 3.1 Hz). Ϫ 1H NMR (250 MHz, [D6]DMSO): acetate. The filtrates were concentrated in a rotary evaporator, and
δ ϭ 1.00 (dd, J ϭ 11.0, J ϭ 3.4, 1 H), 1.32 (m, 1 H), 1.60 (d, J ϭ
8.6 Hz, 1 H), 2.25 (dd, J ϭ 12.0, J ϭ 3.8, 1 H). 2.80 (br. s, 1 H), less solid (0.20 g, 97%) with m.p. 103Ϫ104 °C was obtained. The
2.93 (br. s, 1 H), 6.03 (dd, J ϭ 5.7 Hz, 1 H), 6.30 (dd, J ϭ 5.7, J ϭ
1H NMR spectrum was in agreement with the one obtained from
3.0, 1 H). Ϫ 13C NMR (75.47 MHz, CDCl3): δ ϭ 42.8, 42.9, 49.35, 16a described above.
53.1, 81.6, 132.8, 142.0, 178.1.
13C NMR (62.9 MHz,
the residue was crystallised from 12 mL of PE/EA (5:1). A colour-
Ϫ
2-Hydroxy-1,4-dimethyl-7-oxabicyclo[2.2.1]heptane-2-carboxylic
Acid (23bx): This compound was prepared from 16b (2.23 g,
10.0 mmol) in THF (10 mL) and NaOH (6.04 g, 151 mmol) in
water (53.5 mL). After removal of the solvent, the pale yellow res-
idue was sublimed at 100 °C/0.05 Torr to give 1.11 g (60%) of col-
ourless 23bx, as a solid with m.p. 122Ϫ123 °C. Ϫ IR (KBr): ν˜ ϭ
3300 (broad, OH), 1720 cmϪ1 (COOH). Ϫ 1H NMR (80 MHz,
CDCl3): δ ϭ 1.39 (s, 3 H, 4-CH3), 1.50 (s, 3 H, 1-CH3), 1.58Ϫ2.80
(m, 6 H, 3-H, 5-H and 6-H), 6.24Ϫ7.61 (broad, flat signal, COOH).
[D6]DMSO): δ ϭ 40.1, 41.7, 47.8, 51.1, 80.2, 133.9 138.0, 177.5. Ϫ
EIMS (70 eV): m/z (%) ϭ 154 (2) [Mϩ from C8H10O3], 109 (2) [Mϩ
Ϫ CO2H], 67 (7), 66 (100). Ϫ HRMS: calcd. for C8H10O3 154.0630;
found 154.0632.
2-Hydroxy-7-oxabicyclo[2.2.1]hept-5-ene-2-carboxylic Acid (22x): A
solution of sodium hydroxide (6.04 g, 151 mmol) in water (51 mL)
was added dropwise over 15 min to a solution of 4nn (1.93 g,
10.0 mmol) in THF (8 mL) under nitrogen, with cooling in an ice
bath. The mixture, which turned dark red, was stirred for 3.5 h at
0 °C. The ice bath was removed, and stirring was continued for a
further 1.5 h at room temperature. The mixture was extracted with
ethyl acetate (4 ϫ 20 mL). The aqueous layer was carefully acidified
to pH 1 (indicator paper) in the ice bath with conc. hydrochloric
acid. The acidic solution was extracted with ethyl acetate (8 ϫ
20 mL). After drying with sodium sulfate, the organic extracts were
concentrated in a rotary evaporator to give 1.39 g of a brown oil.
The aqueous layer was further extracted by liquid-liquid continu-
ous extraction with ethyl acetate for 12 h. This operation gave a
further 400 mg of a brown oil. Both oily residues were combined
and dissolved in boiling diethyl ether (50 mL). On cooling, 0.30 g
of a pale yellow solid with m.p. 105Ϫ106 °C (dec.) crystallized.
Ϫ
13C NMR/off-resonance (75.47 MHz, [D6]DMSO): δ ϭ 16.95
(q, 1-CH3), 21.3 (q, 4-CH3), 32.4 (t), 36.8 (t), 49.5 (t), 82.45 (s),
82.7 (s), 87.5 (s), 176.1 (s, COOH). Ϫ EIMS (20 eV): m/z (%) ϭ
186 (19) [Mϩ], 168 (73) [Mϩ Ϫ H2O], 141 (65), 140 (54), 128 (70),
125 (13), 124 (32), 123 (13), 112 (54), 111 (66), 109 (56), 98 (100).
Ϫ C9H14O4 (186.2): calcd. C 58.05, H 7.58; found C 57.83, H 7.50.
2-Chloro-4-hydroxy-1,5-dimethyl-8-oxabicyclo[3.2.1]octan-3-one
(19b): A solution of NaOH (4.06 g, 102 mmol) in water (34 mL)
was rapidly added dropwise, under a nitrogen atmosphere, to a so-
lution of 16b (1.50 g, 6.72 mmol) in THF (7 mL), cooled in an ice
bath. After 4 h stirring with ice cooling, the mixture was neutralised
with conc. hydrochloric acid (to pH 7). The mixture was extracted
with ethyl acetate (5 ϫ 30 mL), and the combined extracts were
dried with magnesium sulfate. After concentration in a rotary evap-
orator, the residue was distilled in a kugelrohr apparatus at 140 °C/
0.005 Torr. The distillate, a pale yellow oil (0.97 g), was dissolved
in 40 mL of boiling PE (40Ϫ65 °C). After cooling, a colourless
solid (0.51 g, 37%) with m.p. 103Ϫ106 °C was obtained. Ϫ IR
˜
Yield 19% 22x. Ϫ IR (KBr): ν ϭ 3420 (sharp, OH), 2960 (broad
absorption, COOH), 1705 cmϪ1 (broad, COOH). Ϫ 1H NMR
(300 MHz, [D6]DMSO): δ ϭ 1.14 (d, J ϭ 11.6 Hz, 1 H, 3n-H), 2.38
(dd, J ϭ 11.6, J ϭ 4.9, 1 H, 3x-H), 4.82 (d, J ϭ 1.2 Hz, the highfield
part is split with J ϭ 0.7 Hz, 1 H, 1-H), 4.91 (d, J ϭ 4.4 Hz, the
lowfield part is split into doublets with J ϭ 0.5 Hz, 1 H, 4-H), AB
sub-spectrum centred at δ ϭ 6.45 with δA ϭ 6.56 and δB ϭ 6.34,
Ϫ1
1
˜
(KBr): ν ϭ 3400 (broad, OH), 1730 cm (COOH). Ϫ H NMR
(80 MHz, CDCl3): δ ϭ 1.38Ϫ2.15 (m, 10 H, surmounted by 2 sing-
lets at 1.46 and 1.58, 6-H, 7-H, 5-CH3 and 1-CH3), 3.45Ϫ3.75 (m,
1 H, OH), 3.83 (s, 1 H, 4-H), 4.79 (d, J ϭ 2.0 Hz, 1 H, 2-H). Ϫ
13C NMR (75.47 MHz, CDCl3): δ ϭ 20.7, 24.1, 30.9, 33.1, 67.9,
81.0, 84.1, 87.3, 199.0 (C-3). Ϫ C9H13ClO3 (204.7): calcd. C 52.82,
H 6.40, Cl 17.32; found C 52.79, H 6.41 Cl 17.49.
J
AB ϭ 5.7 Hz; the lines of the A and B parts are split into doublets
with J ϭ 1.7 Hz, 5-H and 6-H). Ϫ 13C NMR/off-resonance
(75.47 MHz, [D6]DMSO): δ ϭ 175.6 (s), 137.6 (d), 133.6 (d), 81.7
(d), 78.5 (d), 77.9 (s), 38.9 (t). Ϫ C7H8O4 (156.1): calcd. C 53.85,
H 5.16; found C 53.78, H 5.17.
2-Hydroxy-7-oxabicyclo[2.2.1]heptane-2-carboxylic Acid (23ax):
This compound was prepared from 16a (1.95 g, 10.0 mmol) in THF
(8 mL) and NaOH (6.04 g, 151 mmol) in water (51 mL). After
evaporation of the solvent, the viscous residue was dissolved in the
minimum amount possible of boiling PE/EA (5:1). On standing,
pale yellow crystals (0.59 g, 37%) with m.p. 100Ϫ101 °C were col-
lected. A sample for analysis was sublimed at 100 °C/0.03 Torr.
Colourless solid with m.p. 102Ϫ103 °C. Ϫ IR (KBr): ν˜ ϭ 3400,
3200 (broad absorption, OH), 1725 cmϪ1 (COOH). Ϫ 1H NMR
(300 MHz, [D6]DMSO): δ ϭ 1.26 (d, J ϭ 12.4 Hz, 1 H, 3n-H),
1.36Ϫ1.60 (m, 3 H, 3x-H, 5-H, 6n-H), 2.12Ϫ2.20 (m, 1 H, 5-H or
6x-H), 2.29Ϫ2.35 (m, 1 H, 5-H or 6x-H), 4.42Ϫ4.45 (m, 2 H, 1-H,
4-H), 5.72 (broad s, 1 H, HO-), 12.42 (broad s, 1 H, COOH). Ϫ
13C NMR/off-resonance (75.47 MHz, [D6]DMSO): δ ϭ 22.6 (t),
29.15 (t), 42.3 (t, C-3), 76.5 (d), 79.6 (s, C-2), 80.6 (d), 175.5 (s,
COOH). Ϫ C7H10O4 (158.2): calcd. C 53.16, H 6.37; found C 53.01,
H 6.37.
Acknowledgments
We are grateful to Wacker Chemie (Munich) for a gift of 1,1,3-
trichloropropan-2-one (TCA). We thank the analytical and spectro-
scopic service of the Institute of Organic Chemistry for gas chro-
matograms, spectra, and elemental analyses, especially Jochen
Rebell (NMR), and Dr. Wolfgang Frey for help with the X-ray
data. Manfred Flick made an experimental contribution.
[1]
For reviews see: [1a] R. Noyori, Y. Hayakawa, Org. React. 1983,
[1b]
29, 163Ϫ344. Ϫ
H. M. R. Hoffmann, Angew. Chem. 1984,
[1c]
96, 29Ϫ48; Angew. Chem. Int. Ed. Engl. 1984, 23, 1. Ϫ
J.
Mann, Tetrahedron 1986, 42, 4611Ϫ4659. Ϫ [1d] J. H. Rigby, F.
[1e]
C. Pigge, Org. React. 1997, 51, 351Ϫ478. Ϫ
Tetrahedron 1997, 53, 6235Ϫ6280.
H. Kim, H. M. R. Hoffmann, Eur. J. Org. Chem. 2000,
2195Ϫ2201.
B. Föhlisch, D. Krimmer, E. Gehrlach, D. Käshammer, Chem.
Ber. 1988, 121, 1585Ϫ1593.
N. De Kimpe, R. Verhe, The Chemistry of α-Haloketones, α-
M. Harmata,
[2]
[3]
[4]
2-Hydroxy-7-oxabicyclo[2.2.1]heptane-2-carboxylic Acid (23ax) by
Hydrogenation of 22x: Palladium on charcoal (10%) hydrogenation
catalyst (40 mg) was added to a solution of 22x (0.20 g, 1.30 mmol)
in ethyl acetate (25 mL). The mixture was shaken in an atmosphere
´
Haloaldehydes and α-Haloimines, in Updates from the Chem-
4364
Eur. J. Org. Chem. 2001, 4357Ϫ4365

Base-Induced Solvolyses of [3.2.1]Bicyclic α,αЈ-Dichloro Ketones
FULL PAPER
istry of Functional Groups (Eds.: S. Patai, Z. Rapoport), Wiley,
Chichester, 1988.
A. Baretta, B. Waegell, Tetrahedron Lett. 1976, 753Ϫ756.
P. J. Chenier, J. C. Kao, J. Org. Chem. 1976, 41, 3730Ϫ3734.
MO calculations on the ions 9 and 13, and also its O-pro-
tonated equivalents, would be desirable to give further insights
into the competing reaction channels affording 10, 11, and 14.
To simplify matters in Scheme 1, intermediates are formulated
as species with negatively charged oxygen atoms.
S. Sendelbach, R. Schwetzler-Raschke, A. Radl, R. Kaiser, G.
H. Henle, H. Korfant, S. Reiner, B. Föhlisch, J. Org. Chem.
1999, 64, 3398Ϫ3408.
O. Wallach, Justus Liebigs Ann. Chem. 1918, 413, 296Ϫ366.
O. Wallach, A. Weißenborn, Justus Liebigs Ann. Chem. 1924,
437, 148Ϫ186.
See, e.g.: S. Selman, J. F. Eastham, Quart. Rev. Chem. Soc.
1960, 14, 221Ϫ235.
M. Utaka, S. Matsushita, A. Takeda, Chem. Lett. 1980,
779Ϫ780.
Cambridge CB2 1EZ, UK [Fax int. code ϩ 44(1223)336-033;
E-mail: deposit@ccdc.cam.ac.uk].
[18]
[5]
In the latter case the negative charge of the oxyanion interme-
diate must be put on the ‘‘left’’ oxygen atom of the interme-
diate. However that would be an irrelevant formalism since
rapid prototropy would be established in water. Ϫ A SET
mechanism might also be considered; for MO calculations see:
I. Rajyaguru, H. S. Rzepa, J. Chem. Soc., Perkin Trans. 2
1987, 1819Ϫ1827.
[6]
[7]
[19]
[8]
However, under forced conditions, cyclohexane-1,2-dione also
undergoes rearrangement to 1-hydroxycyclopentanecarboxylic
acid [ref. 10].
[20] [20a]
[20b]
A. S. Kende, Org. React. 1960, 11, 261Ϫ316. Ϫ
A. A.
[9]
Akhrem, T. K. Ustynyuk, Y. A. Titov, Russ. Chem. Rev. 1970,
39, 732Ϫ746. Ϫ [20c] A. Baretta, B. Waegell, A Survey of Favor-
skii Rearrangement Mechanisms, in: Reactive Intermediates
(Ed.: R. A. Abramovitch), vol. 2, p. 527Ϫ585, Plenum Press,
New York Ϫ London, 1982.
[10]
[11]
[12]
[21]
[22]
However, α-bromocyclobutanone is an exception to this rule:
[21a]
J. M. Conia, J. L. Ripoll, Bull. Soc. Chim. Fr. 1963,
[13]
K. Alder, R. Hartmann, W. Roth, Chem. Ber. 1960, 93,
2271Ϫ2281.
K. Alder, R. Hartmann, W. Roth, Justus Liebigs Ann. Chem.
[21b]
755Ϫ767. Ϫ
J. M. Conia, J. Salaun, Bull. Soc. Chim. Fr.
[21c]
1964, 1957Ϫ1963. Ϫ
C. Rappe, L. Knutsson, Acta Chem.
[14]
Scand. 1967, 21, 163Ϫ167. Ϫ [21d] R. Castillo, V. Moliner, V. S.
1958, 613, 6Ϫ27.
´
Safont, M. Oliva, J. Andres, Theochem 1998, 426, 299Ϫ306.
[15] [15a]
E. E. van Tamelen, M. Shamma, J. Am. Chem. Soc. 1954,
The careful reader will note that the products formed from
meso dichlorides are chiral molecules; might the reactions be
executed with enantioselective control?
The authors will not continue these investigations and therefore
would welcome further research.
[15b]
76, 2315Ϫ2317. Ϫ
C. D. Ver Nooy, Ch. S. Rondestvedt,
Jr., J. Am. Chem. Soc. 1955, 77, 3583Ϫ3586.
For details see: A. Radl, Dissertation, University of Stuttgart,
1992.
Crystallographic data (excluding structure factors) for the
structure reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary
publication no. CCDC-157688. Copies of the data can be ob-
tained free of charge on application to CCDC, 12 Union Road,
[16]
[17]
[23]
[24]
Prepared with o-phenylenediamine in analogy to the derivative
of bornane-2,3-dione (3-oxocamphor): A. Heckendorn, Helv.
Chim. Acta 1929, 12, 50Ϫ60.
Received April 20, 2001
[O01190]
Eur. J. Org. Chem. 2001, 4357Ϫ4365
4365