Photocycloadditions of tetrachloro-1,4-benzoquinone (chloranil) onto cyclobutene and cyclopropene. Expected and unexpected products

Article abstract of DOI:10.1039/c2ob07048k

Solutions of chloranil (CA) in chlorobenzene were irradiated in the presence of cyclobutene and cyclopropene. Cyclobutene gave rise to two conventional 1:2 cycloadducts onto the dichloroethene subunits of CA and an α,β-unsaturated α,γ-dichloro-γ-lactone. Heating of the crude product in methanol converted the lactone into an α,β- unsaturated methyl γ-oxocarboxylate (25% yield) and a large amount of the major 1:2 cycloadduct, which contains chlorocyclobutane entities, into a cyclopropylcarbinyl chloride derivative (24% yield). An entirely new product type was the result in the case of cyclopropene. After treatment of the crude product with methanol a tetracyclic acetal containing a cyclopentanone and a dihydropyran subunit was isolated in 36% yield. Apparently, CA had taken up two molecules of cyclopropene. One of the resulting cyclopropane entities must have undergone a rearrangement en route to the final product.

Full text of DOI:10.1039/c2ob07048k

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 4400
www.rsc.org/obc
PAPER
Photocycloadditions of tetrachloro-1,4-benzoquinone (chloranil) onto
cyclobutene and cyclopropene. Expected and unexpected products†
Max Braun and Manfred Christl*
Received 7th December 2011, Accepted 11th April 2012
DOI: 10.1039/c2ob07048k
Solutions of chloranil (CA) in chlorobenzene were irradiated in the presence of cyclobutene and
cyclopropene. Cyclobutene gave rise to two conventional 1 : 2 cycloadducts onto the dichloroethene
subunits of CA and an α,β-unsaturated α,γ-dichloro-γ-lactone. Heating of the crude product in methanol
converted the lactone into an α,β-unsaturated methyl γ-oxocarboxylate (25% yield) and a large amount of
the major 1 : 2 cycloadduct, which contains chlorocyclobutane entities, into a cyclopropylcarbinyl
chloride derivative (24% yield). An entirely new product type was the result in the case of cyclopropene.
After treatment of the crude product with methanol a tetracyclic acetal containing a cyclopentanone and a
dihydropyran subunit was isolated in 36% yield. Apparently, CA had taken up two molecules of
cyclopropene. One of the resulting cyclopropane entities must have undergone a rearrangement en route
to the final product.
formation of the tetrachlorobenzofuranone derivative. The
Introduction
results with the more strained cyclopentenes norbornene and
bicyclo[2.1.1]hex-2-ene3b support this assumption. On the other
Reactions of quinones with alkenes have been an area of con-
tinuous interest in organic photochemistry.1 Those of chloranil
hand, trans-cyclooctene does not bring about a tetrachlorobenzo-
(CA) are notable because of the great diversity of possible pro-
cesses. Owing to the rapid intersystem crossing, all conversions
of excited CA that lead to chemical change of CA start from the
furanone in spite of its relatively high strain energy of 15.3 kcal
mol−1,5 but only a [2 + 2] cycloadduct onto a dichloroethene
subunit of CA.3a The reason could be the different origin of the
strain in comparison to the situation in cyclopentene, namely the
pyramidalisation of the ethene moiety, which may not promote
the unusual reactivity. To examine whether or not an enhanced
angle strain relative to that of cyclopentene will increase the
portion of the tetrachlorobenzofuranone product, we irradiated
CA in the presence of cyclobutene and cyclopropene. With these
3
triplet state (3CA).2 The most important reaction types of CA
with alkenes have been summarised in the introduction of our
recent publication.3a Here we restrict ourselves to a brief review
of results pertinent to the reactions described in this work, that
is, to the three processes that a monocyclic monoalkene may
3
undergo with CA. Whereas cis- and trans-cyclooctene exclu-
3
sively undergo a [2 + 2] cycloaddition onto a dichloroethene
subunit,3a cyclohexene, dissolved in benzene, further gives rise
to the photoreduction of chloranil (CA) with formation of tetra-
chlorohydroquinone and its monocyclohexenyl ether.3a,4 In
addition to the generation of two dichlorocyclobutane derivatives
and the photoreduction, a third reaction type is exhibited by
cyclopentene, as a tetrachlorobenzofuranone derivative results.
En route to the latter product, probably a [4 + 2] cycloaddition
cycloalkenes, an electron transfer onto CA, as it proceeds or
can be assumed with substrates of a low oxidation potential,3c,4,6
was not expected, since the estimation of the oxidation potentials
on the basis of first the vertical ionisation potentials7 (cyclobu-
tene: 9.43,8 9.59 eV;9 cyclopropene: 9.86,8 9.82 eV9) gives
values considerably higher than those of cyclopentene and
cyclohexene (first vertical ionisation potentials: 9.18 eV and
9.12 eV,8 respectively). And in the latter cases, at most the photo-
reduction of CA can be considered as a process initiated by an
electron transfer from the cycloalkene onto 3CA.4
3
proceeds, in which CA serves as substrate with 4 π electrons.3a,
b
Even if the angle strain of the ethene moiety of cyclopentene
is only very small, it could be the deciding factor for the
Results and discussion
At variance with the reactions recently described by us,3a
benzene could not be utilised as solvent, because the irradiations
of CA in the presence of cyclobutene and cyclopropene should
be carried out at temperatures below 0 °C due to the volatility of
Institute of Organic Chemistry, University of Würzburg, Am Hubland,
D-97074, Germany. E-mail: christl@chemie.uni-wuerzburg.de;
Fax: +49 931 3184606; Tel: +49 931 3185344
†Dedicated to Professor Waldemar Adam on the occasion of his 75th
birthday.
4400 | Org. Biomol. Chem., 2012, 10, 4400–4406
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 1 Compounds isolated after the photochemical reaction of
chloranil (CA) with cyclobutene and the treatment of the crude product
with methanol.
3
these cycloalkenes. Since toluene reacts readily with CA,3c it
could not be used, which is why we chose chlorobenzene.
1. Cyclobutene
The reaction was performed at −5 °C with a medium-pressure
mercury lamp by utilising a pyrex immersion well containing a
glass filter, which surrounded the lamp and was supposed to
prevent the passage of light of λ < 400 nm. After the complete
consumption of CA, chlorobenzene was evaporated at 35 °C/
15 mmHg and the residue refluxed in methanol. In this way, a
possibly formed tetrachlorobenzofuranone derivative, which
should be a pseudoacid chloride, would be transformed to a
pseudoester or an ester.3b
Indeed, after the evaporation of the methanol, the methyl ester
1 was isolated by flash chromatography in 25% yield. In addition
to it, the 1 : 2 cycloadducts 2 and 3 of CA as well as their
unsymmetrical isomer 4 (Scheme 1) were obtained in 6, 10 and
24% yield, respectively. The identity of these products is based
on their NMR spectroscopic data, and there the comparison with
the values of analogous products plays an important part. In the
case of 1, these are the corresponding methyl esters derived from
norbornene and cyclopentene3b and with 2 and 3 the 1 : 2
cycloadducts of cyclopentene,3b cyclohexene and cis-
cyclooctene.3a
Scheme 2 The mechanism for the formation of the compounds on
irradiation of chloranil (CA) in the presence of cyclobutene followed by
heating of the crude product in methanol.
from a bicyclo[2.2.0]hexane moiety. The parent hydrocarbons
are different by 20.8 kcal mol−1 with respect to their standard
heat of formation.10
In the NMR spectra of the crude product prior to methanoly-
sis, only the signals of 3 and another compound were clearly dis-
cernible. We take the latter for the tetrachlorobenzofuranone
derivative 6 (Scheme 2), which should have yielded the ester 1
on treatment with methanol. The ratio of 3 : 6 amounted to
1.4 : 1. That the absorptions of the 1 : 2 cycloadduct 2 could at
best be suspected, was probably caused by the small quantity
(6% isolated yield), which gave rise only to an insufficient
signal-to-noise ratio. Although 3 was isolated in a small yield as
well (10%), its bands appeared with a good signal-to-noise ratio
in the spectra of the crude product. This is ample evidence for
the conversion of 3 into 4 to a substantial extent on heating in
methanol. This rearrangement takes advantage from the decrease
of the ring strain energy, since a bicyclo[3.1.0]hexane emerges
In all probability, this isomerisation proceeded by the heteroly-
tic dissociation of a C–Cl bond of 3 with concomitant Wagner–
Meerwein rearrangement furnishing the contact ion pair 9
(Scheme 2), which is a derivative of the cyclopropylcarbinyl
cation. The collapse of 9 then led to the isolated compound 4,
the possible configurations and conformations of which are dis-
cussed below. In spite of the presence of the nucleophile metha-
nol, only the rearrangement product 4 was detected and no
methanolysis product. However, a small amount of the methyl
ether corresponding to 4 could have been overlooked in the
workup.
The reverse isomerisation, that is, the formation of cyclobutyl
chloride from cyclopropylcarbinyl chloride on solvolysis of the
latter had been observed a long time ago.11 A more recent
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 4400–4406 | 4401

View Article Online
example for internal return after the reorganisation of the carbon
atom skeleton is the generation of endo-2-norbornyl mesylate as
major product of ethanolysis and hydrolysis of 2-norpinyl
mesylate.12
The formation of the tetrachlorobenzofuranone derivative 6
should have proceeded on a route analogous to that proposed for
the production of the corresponding compounds from norbor-
nene, bicyclo[2.1.1]hex-2-ene and cyclopentene.3a,b Thus, cyclo-
3
butene would have undergone a [4 + 2] cycloaddition with CA
and afforded the triplet diradical 7, which would have been
subject to opening of the ring originating from CA after intersys-
tem crossing to give the α-chloro-β-oxoketene derivative 5
(Scheme 2). The closure of a five-membered ring by addition of
the ketone oxygen atom onto the central ketene carbon atom of 5
and the [1,2] migration of the α chlorine atom then could have
completed the pathway to 6.
In comparison to the reaction of cyclopentene, the percentage
of the [4 + 2] cycloaddition of cyclobutene onto 3CA is not
increased. Thereby, the argument advanced in the introduction
that this reaction type could be supported by the angle strain of
the ethene subunit of cyclopentene, norbornene and bicyclo
[2.1.1]hex-2-ene is not invalidated, however, since this quantity
of cyclobutene is not much higher than that of cyclopentene.
This can be estimated from the total strain energies of cyclopen-
tene and cyclopentane (5.9 and 6.3 kcal mol−1) on the one hand
and cyclobutene and cyclobutane (29.8 and 26.2 kcal mol−1) on
the other.5
Fig. 1 Possible configurations of the photocycloadducts 2 and 3 of CA
onto cyclobutene and of the product 4 resulting from the rearrangement
of 3.
Though we speculated in an earlier paper3b that the [4 + 2]
Possibly, the reaction of excited 8a and/or 8b with cyclobutene
proceeds even faster than that of 3CA.
3
cycloaddition of CA onto cyclopentene and its derivatives was
With 8a and 8b being the possible precursors, the diastereo-
mers 2a and 2b have to be considered as the Cs symmetric 1 : 2
cycloadduct (Fig. 1). The difference between the two consists in
the relative arrangement of the bicyclo[2.2.0]hexane moieties at
the cyclohexanedione core, which is trans in 2a and cis in 2b.
The formation of 2b could be at a disadvantage with that of 2a,
since cyclobutene would have to approach the excited 8a or 8b
from the endo side, which is sterically strongly hindered by the
bicyclo[2.2.0]hexane part. The configuration of 2a is analogous
to that of the sole 1 : 2 cycloadduct of cyclopentene, which has
been analysed by X-ray diffraction.3b
In the case of the C2h or C2v symmetric 1 : 2 cycloadduct,
the four diastereomers 3a–3d are possible (Fig. 1). As with the
comparison of 2a and 2b, we favour 3a and 3b because of the
trans annulation of the bicyclo[2.2.0]hexane systems at the
cyclohexanedione core over 3c and 3d, although there is no
experimental evidence for this assumption. However, the most
likely configuration of the rearrangement product 4 puts stereo-
chemical restrictions on the precursor 3 (see below).
concerted, the mechanism has not yet been shown and the dis-
tinction between concerted and two step radical processes has
not yet been proved. Cyclobutene would support a concerted
reaction, since it adds onto dimethyl 1,2,4,5-tetrazine-3,6-dicar-
boxylate in a concerted [4 + 2] cycloaddition 14 times as fast as
cyclopentene and still with one third of the rate of norbornene.13
Also, cyclohexene and cis-cyclooctene, which react significantly
slower than cyclopentene with the tetrazine,13 would not militate
3
against concertedness of the [4 + 2] cycloaddition onto CA, as
they furnish only [2 + 2] cycloadducts and no product of the [4
+ 2] mode.3a However, trans-cyclooctene severely questions
3
concertedness of the [4 + 2] cycloaddition onto CA, since it
combines over 1000 times as fast as norbornene with the tetra-
zine13 and yet does not give a product formed via a [4 + 2]
cycloaddition onto 3CA.3a
The stereochemical alternatives of the 1 : 2 cycloadducts 2 and
3 are the same as in the case of cyclopentene as substrate for
3CA. At variance with the reaction of cyclohexene,3a an antara-
facial cycloaddition can be excluded on the grounds of ring
strain of the potential product. Thus, as conceivable precursors
for 2 and 3, the 1 : 1 cycloadducts 8a and 8b (Scheme 2) have to
be discussed, in which the six-membered and the terminal four-
membered ring are located trans and cis, respectively, at the
central cyclobutane subunit. However, we observed neither 8a
nor 8b. Apparently, the excitation of these compounds by the
light used and the subsequent reaction with the excess of cyclo-
butene occurred more readily than the corresponding processes
of the 1 : 1 cycloadducts of CA with cyclopentene, cyclohexene
and cis-cyclooctene under similar conditions. There, 1 : 1
cycloadducts were observed and some of them even isolated.3a,b
The 1H NMR spectrum of 4 contains characteristic infor-
mation on the steric arrangement of the bicyclo[3.1.0]hexane
system. Causing a doublet with the coupling constant of 5.4 Hz,
the proton at the three-membered ring, therefore, interacts only
with the exo proton of the proximal methylene group. This pattern
is typical for the boat conformation of the bicyclo[3.1.0]hexane
skeleton,14 which has been known for the parent hydrocarbon
for a long time.15,16 The multiplicity of the signal of the CHCl
group is revealing as well, as it is a triplet with a line distance of
9.2 Hz. Thus, this proton is coupled equally strongly with both
protons of the neighbouring methylene group. This situation is
4402 | Org. Biomol. Chem., 2012, 10, 4400–4406
This journal is © The Royal Society of Chemistry 2012

View Article Online
in accord only with the exo position of this proton at the bicyclo
[3.1.0]hexane system, since, as an endo proton, it would exhibit
a resolved interaction exclusively with the endo proton of the
neighbouring methylene group.14
Based on these findings, configuration and conformation of
the bicyclo[3.1.0]hexane subunit of 4 are determined as illus-
trated by the formulas 4a and 4b in Fig. 1. The chlorine atom at
the cyclopropane moiety must be located endo, since, owing to
the carbonyl groups and the annulation of two small rings, the
six-membered ring is rather rigid and cannot be annulated trans
at the cyclopropane moiety due to intolerable ring strain. In con-
sequence, chlorine and hydrogen atom at the cyclopropane entity
must be arranged trans. If it is further taken into account that the
configuration of the migrating centre is retained in the Wagner–
Meerwein rearrangement, only those isomers 3 have to be con-
sidered as precursors of 4 that have a trans relationship between
a chlorine atom and the hydrogen atom next to it, that is, 3a and
3c. In contrast, the rearrangement of 3b and 3d would lead to
the trans annulation of the five- or six-membered ring at the
cyclopropane moiety. As discussed above, we favour 3a and,
hence, 4a over 3c and 4b, respectively.
Scheme 3 The compound 10 isolated after the photochemical reaction
of chloranil (CA) with cyclopropene and the treatment of the crude
product with methanol.
Our imagination of the reorganisation of 3a or 3c starts with
an inspection of the bond common to the two cyclobutane sub-
units of a bicyclo[2.2.0]hexane system. Because of the angle
strain, the electron density is bent out of the straight line
between the respective carbon nuclei on the side opposite to
the chlorine atoms. This electron pair should have attacked a
chlorine-bearing carbon atom and displaced the chlorine atom
from the rear, resulting in the cyclopropylcarbinyl cation within
the ion pair 9 (Scheme 2). Without the necessity to change the
face of the ring system, the chloride ion now should have been
bound by the carbocationic centre giving rise to 4a or 4b
(Fig. 1). Overall, the migrating CH group and the chloride ion,
hence, would have undergone a [1,2] transposition. The present
rearrangement is closely related to the conversion of endo-6-
chlorobicyclo[3.1.0]hexane into 3-chlorocyclohexene.17
2. Cyclopropene
Scheme 4 Mechanistic proposals for the photochemical reaction of
chloranil (CA) with cyclopropene and the methanolysis of the initial
product.
The irradiation of CA in the presence of cyclopropene was
carried out under the same conditions as in the case of cyclobu-
tene, except that now the temperature was kept at −30 °C. After
the complete consumption of CA, the major amount of the
solvent chlorobenzene was quickly evaporated at 35 °C/
15 mmHg, the residue immediately dissolved in methanol and
the mixture left at 20 °C for 24 hours. By flash chromatography,
the tetracyclic acetal 10 (Scheme 3) was then isolated as the sole
product in 36% yield. Its structure was determined by X-ray
diffraction.18
As the skeleton of 10 testifies, CA took up two cyclopropene
molecules by the excitation. One of them retained the three-
membered ring, whereas the other, with inclusion of a chlorine-
bearing carbon atom, underwent the rearrangement to a homoal-
lyl system. Such a reorganisation is typical for cyclopropylcarbi-
nyl subunits carrying a leaving group. In the first section, we
have quoted the transformation of the unsubstituted cyclopropyl-
carbinyl chloride to chlorocyclobutane, parallel to which homo-
allyl chloride is generated.11 In this way, the homoallyl chloride
14 should have been formed from the cyclopropylcarbinyl
chloride 13, namely by heterolytic dissociation and opening of
the three-membered ring giving rise to the ion pair 12
(Scheme 4), the collapse of which should have resulted in 14.
The conversion of the α-chloro ether 14 into the acetal 10 is con-
sidered to be an SN1 substitution with the ion pair 12 again
being the intermediate. The ease of the isomerisation 13 → 14,
The 1H NMR spectrum of the crude product prior to methano-
lysis indicated a complex mixture, whose main component we
take for the α-chloro ether 14 (Scheme 4), which should have
been the precursor to 10. In particular, chemical shifts and coup-
ling constants of four signals are very similar to those of the
cyclopropane protons of 10. In addition, the multiplets at δ 6.24
and 6.39 should originate from the olefinic proton and the CHCl
group. Only the bands of the allylic methylene group cannot be
identified unambiguously.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 4400–4406 | 4403

View Article Online
which did occur already under the conditions of the photochemi-
cal reaction (−35 °C) or the workup (up to 35 °C), has probably
its cause in the oxygen atom bound to the cyclopropane moiety
and being a good electron donor. In consequence, the cation of
12 is a carbenium–oxonium ion and thus highly stabilised.
To rationalise the formation of 13, we propose the diradicals
11a and 11b, which should have been generated by addition of
one cyclopropene molecule onto 3CA and should have furnished
13 by reaction with a second one. Thus, a single excitation of
CA would have led to the taking up of two cyclopropene mol-
ecules. Owing to the triplet nature of the reacting CA,2 an inter-
mediate 11a or 11b should have resulted as a triplet diradical. En
route to 13, an intersystem crossing would then have been
required.
cycloalkenes cyclopentene, cyclohexene and cyclooctene, such
a transformation was not observed.
An entirely new reaction type was found by irradiation of CA
in the presence of cyclopropene. After the methanolysis of the
crude product, the tetracyclic acetal 10 was isolated. It is
believed that two subsequent [3 + 2] cycloadditions of cyclopro-
3
pene onto CA furnished the pentacyclic intermediate 13, which
underwent a ring expansion to give the 2-chlorodihydropyran
14. Dissolved in methanol, the latter was subject to an SN1 sub-
stitution with formation of 10.
Because of the product mixture, the photocycloaddition of
cyclobutene onto CA will probably not gain preparative impor-
tance. However, the single product obtained in the case of cyclo-
propene could be of interest as starting material for syntheses, in
particular, as cyclopropene is easily accessible in quantity.25
Additionally, a number of cyclopropene derivatives are readily
prepared in useful amounts26 and can be envisaged for the syn-
thesis of substituted derivatives of 10.
Both 11a and 11b are derivatives of the oxyallyl diradical. In
the case of 11b, a vinyl group extends the π-electron system.
Recently, the unsubstituted oxyallyl diradical was observed spec-
troscopically for the first time. It is characterised by only a small
energy gap between the singlet and the triplet state, with the
The most interesting aspect of this work and the previous
former being the more stable one by 1.3 kcal mol−1 19
Also, this
.
studies on this subject3a,b,27 is the puzzling reactivity of 3CA
3
finding was compared to the results of previous and the most
with alkenes that do not transfer a single electron to CA. The
recent theoretical studies.19,20
[2 + 2] cycloaddition onto a dichloroethene subunit proceeds
most frequently. In addition to that, the [4 + 2] cycloaddition
The resonance structures of 11a and 11b in Scheme 4 have
been chosen such that they illustrate the formation of 13 by a
[3 + 2] cycloaddition each with cyclopropene. The reaction of
3CA with cyclopropene to give 11a or 11b is also a [3 + 2]
3
involving the positions 2 and 5 of CA takes place with cyclo-
pentenes and cyclobutene and is the major reaction by far in the
case of norbornene. Finally, a [3 + 2] cycloaddition onto either
the positions 2 and 6 or an oxygen and a proximal chlorine-
bearing carbon atom of 3CA was exclusively observed in the
case of cyclopropene. It is very possible that these three addition
types start with a common first step, that is, the making of a C–C
bond resulting in a triplet diradical. These species would then
collapse to give rings of different size according to the kind of
alkene. There is an alternative first step conceivable only in the
case of cyclopropene, as a C–O bond could initially be formed
en route to the possible intermediate 11b. Quantum chemical cal-
culations could greatly contribute to an understanding of these
phenomena.
3
cycloaddition. Such processes of CA are unknown heretofore.
Compared with that, [3 + 2] cycloadducts of oxyallyl systems
with 1,3-dienes21 and vinyl ethers22 were obtained and shown to
be of the cyclopentanone and α-methylenetetrahydrofuran types.
A quantum chemical investigation of the reaction between the
uncharged oxyallyl and 1,3-butadiene indicated, for the [3 + 2]
cycloaddition, the route to an α-methylenetetrahydrofuran as the
most favourable pathway.23
It is interesting to note that oxyallyl intermediates can be gen-
erated photochemically from cross-conjugated cyclic dienones.24
This class of compounds is related to quinones, such as CA. In
particular, a transient oxyallyl with a bicyclic skeleton can be
trapped on irradiation of 2,7-cyclooctadienone in the presence
of vinyl ethers to give tricyclic norbornanones and
oxatriquinanes.22,24
Experimental
General details and general conditions for the photochemical
reactions
See ref. 3a.
Conclusions
Irradiation of chloranil (CA) in the presence of cyclobutene
On employment of cyclobutene, we have discovered the first
alkene that is not a derivative of cyclopentene and yet undergoes
a [4 + 2] cycloaddition with 3CA. With opening of the ring orig-
inating from CA, this reaction is believed to yield the tetrachlor-
obenzofuranone derivative 6 and, from it by the action of
methanol, the isolated product, namely the α,β unsaturated
methyl γ-oxocarboxylate 1. In addition to 1, a Cs and a C2h or
C2v symmetric 1 : 2 cycloadduct (2 and 3, respectively) onto the
ethene subunits of CA were obtained as well as the rearrange-
ment product 4 of 3. Most probably promoted by the
polar solvent methanol, this rearrangement brought about
a cyclopropylcarbinyl chloride from a chlorocyclobutane. In
the case of the 1 : 2 cycloadducts of CA with the higher
In a photolysis vessel, a solution of CA (1.00 g, 4.07 mmol) in
anhydrous chlorobenzene (150 cm3) was saturated with nitrogen
and then cooled to −30 °C. Having been stored in a freezer
under nitrogen, cyclobutene28 (440 mg, 8.14 mmol) was added.
The mixture was then irradiated by utilising a pyrex immersion
well containing a Hanovia mercury lamp (medium pressure, 450
W), which was surrounded by a glass filter supposed to prevent
the passage of light of λ < 400 nm, at −5 °C. After CA had been
completely consumed (6 h), the solvent was quickly (15 min)
evaporated at 35 °C (bath)/15 mmHg and the remaining yellow
oil was immediately refluxed in anhydrous methanol (150 cm3)
over 18 h. The mixture was then concentrated in vacuo and the
4404 | Org. Biomol. Chem., 2012, 10, 4400–4406
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 1 Proton–proton coupling constants of 1, with which the
simulation of the multiplets gives an excellent match with the
experimental fine structure
(2 H, m) and 3.67 (2 H, m); δC (63 MHz; CDCl3) 23.4 (CH2),
23.5 (CH2), 43.4 (CH), 47.6 (CH), 68.9 (CCl), 81.3 (CCl) and
192.8 (CO); m/z (EI, 70 eV) 354 (M+, 0.1%), 175 (15), 143
(30), 141 (91), 115 (25), 113 (34), 105 (22), 78 (20), 77 (100),
53 (24), 51 (35) and 39 (22).
Coupling constants (Hz) with the protons at δ
Chemical shifts
(δ/ppm)
2.196
9.8
2.550
2.670
3.560
4.030
Compound 3, colourless crystals, mp 170–171 °C (Found: C,
47.2; H, 3.2. C14H12Cl4O2 requires C, 47.5; H, 3.4%); νmax
(KBr)/cm−1 3010w, 2994w, 2958m, 1731s, 1722s, 1436w,
1266w, 1259m, 1163m, 1139m, 1102m, 1067m, 1054m, 1015m,
942w, 925m, 915w, 810w, 693m and 673m; δH (250 MHz,
CDCl3) 2.42–2.69 (8 H, m) and 3.63 (4 H, m); δC (63 MHz;
C6D6) 22.2 (CH2), 42.2 (CH), 76.9 (CCl) and 192.1 (CO); m/z
(EI, 70 eV) 321, 319, 317 (M+ − Cl, 4, 11, 12%), 267 (22), 266
(20), 265 (55), 263 (60), 239 (26), 238 (22), 237 (69), 235 (73),
233 (22), 231 (34), 149 (20), 143 (24), 141 (52), 115 (20), 113
(24), 105 (23), 77 (79), 75 (23), 54 (100), 51 (41) and 39 (26).
Compound 4, colourless crystals, mp 213–214 °C (Found: C,
47.9; H, 3.2. C14H12Cl4O2 requires C, 47.5; H, 3.4%); νmax
(KBr)/cm−1 2986m, 2948m, 2931m, 2856w, 1726s, 1700s,
1458w, 1320w, 1308w, 1287w, 1260m, 1245w, 1231w, 1209m,
1156m, 1067w, 1058w, 1006w, 946w, 938m, 928m, 907m,
855w, 710m and 616w; δH (400 MHz; CDCl3) 2.28 (1 H, m),
2.36–2.61 (7 H, m), 3.25–3.33 (2 H, m), 3.64 (1 H, d, J 5.4, H
on cyclopropane moiety) and 5.16 (1 H, t, J 9.2, CHCl); δC
(63 MHz; CDCl3) 20.9 (CH2), 21.1 (CH2), 26.2 (CH2), 33.6
(CH2), 38.8 (CH), 40.3 (CH), 44.9 (CH), 58.3 (CHCl), 58.8,
66.2, 69.6 and 70.7 (3 CCl and Cq), 189.6 and 193.5 (2 CO); m/
z (EI, 70 eV) 356, 354, 352 (M+, 0.2, 0.4, 0.3%), 321, 319, 317
(M+ − Cl, 7, 24, 24), 283 (23), 281 (34), 253 (20), 247 (20),
245 (44), 217 (28), 209 (29), 181 (23), 153 (22), 143 (21), 141
(67), 115 (20), 113 (22), 105 (40), 77 (100) and 51 (40).
2.124
2.196
2.550
2.670
3.560
3.1
−11.6
−11.8
9.8
9.4
3.0
−1.1
2.0
−0.9
9.3
9.0
0.0
8.1
9.5
residue subjected to flash chromatography (SiO2, pentane–ethyl
acetate 25 : 1). Four fractions, a yellow oil each, were collected,
from which crystals were obtained after dissolution in the
minimum quantity of ethyl acetate and storage of the solutions at
−30 °C. The order of the elution was: (2aα,2bα,3aβ,3bα,5aα,5b-
β,6aα,6bα)-
2a
or
(2aα,2bβ,3aβ,3bβ,5aβ,5bβ,6aβ,6bα)-
2b,3a,5b,6a- tetrachlorododecahydrodicyclobuta[1,2:1′,2′]dicy-
clobuta[3,4-a:3′,4′-d]benzene-3,6-dione 2b (81 mg, 6%),
(2aS,2bR,3aS,4S,6aR,6bR,7aS,7bR)-rel- 4a or (2aR,2bS,3aS,-
4S,6aR,6bR,7aR,7bS)-rel-2b,4,6b,7a-tetrachlorodecahydro(cyclo-
penta[1,2]cyclopropa)(cyclobuta[1′,2′]cyclobuta)[2,3-a:3′,4′-d]-
benzene-3,7(1H)-dione 4b (348 mg, 24%), methyl (E)-
[(2aα,6aα)-5,6-dichloro-2,2a,4,6a-tetrahydro-4-oxocyclobuta-
benzen-3(1H)-ylidene]chloroacetate
1 (296 mg, 25%) and
(2aα,2bβ,3aα,3bβ,5aβ,5bα,6aβ,6bα)- 3a or (2aα,2bα,3aβ,3b-
β,5aβ,5bβ,6aα,6bα)- 3b or (2aα,2bβ,3aβ,3bα,5aα,5bβ,6aβ,6bα)-
3c or (2aα,2bα,3aα,3bα,5aα,5bα,6aα,6bα)-2b,3a,5b,6a- tetra-
chlorododecahydrodicyclobuta[1,2:1′,2′]dicyclobuta[3,4-a:3′,4′-d]
benzene-3,6-dione 3d (151 mg, 10%).
In the NMR spectra of the crude product prior to methano-
lysis, only the signals of chlorobenzene, 3 and a third component
could be unambiguously discerned. We take the latter for the
pseudoacid chloride 6, which should be converted into the
methyl ester 1 by methanol; δH of the third component
(250 MHz; CDCl3) 1.70–3.80 (several m); δC of the third com-
ponent (63 MHz; CDCl3) 22.0 (CH2), 26.7 (CH2), 32.0 (CH)
and 49.4 (CH); the signals of the carbon atoms without a hydro-
gen atom were not observed due to too low intensities.
Compound 1, colourless solid (the formation of crystals lasted
4 weeks), mp 99–101 °C (Found: C, 45.0; H, 3.1. C11H9Cl3O3
requires C, 44.7; H, 3.1%); νmax (KBr)/cm−1 2994m, 2958m,
2936m, 1732s, 1668s, 1601s, 1581s, 1451m, 1435s, 1334m,
1305s, 1285s, 1268s, 1247m, 1242m, 1211s, 1200m, 1122m,
1050s, 1029m, 970m, 934w, 882m, 852w, 808w, 774w, 751m,
734m, 660m, 642m and 622m; δH (400 MHz; CDCl3) 2.124
(1 H, m) and 2.670 (1 H, m) (CH2), 2.196 (1 H, m) and 2.550
(1 H, m) (CH2), 3.56 (1 H, m, CH), 3.91 (3 H, s, CH3) and 4.03
(1 H, m, CH). The fine structure of the multiplets is distinctive,
but cannot be described in simple terms due to higher order
effects. Therefore, we have simulated the shape of the multiplets
by using the gNMR program.29 The coupling constants listed in
Table 1 furnish an excellent match and, thus, testify to a signifi-
cant folding of the cyclobutane subunit. δC (63 MHz; CDCl3)
26.1 (CH2), 26.3 (CH2), 33.8 (CH), 41.9 (CH), 53.5 (CH3),
130.8 (CCl), 132.2 (CCl), 133.8 (CCl), 156.4 (C-3), 164.6
(CO2CH3) and 175.5 (C-4); m/z (EI, 70 eV) 298, 296, 294 (M+,
5, 16, 17%), 270 (26), 268 (78), 266 (83), 253 (23), 251 (24),
235 (28), 233 (63), 231 (100), 227 (22), 203 (20), 181 (24), 179
(24), 174 (28), 172 (43) and 59 (32).
Irradiation of chloranil (CA) in the presence of cyclopropene
In a photolysis vessel, a solution of CA (800 mg, 3.25 mmol) in
anhydrous chlorobenzene (150 cm3) was saturated with nitrogen
and then cooled to −30 °C. Cyclopropene was prepared by treat-
ment of allyl chloride (23.0 g, 301 mmol) with sodium amide
(12.0 g, 308 mmol) according to Closs and Krantz.25a It was
driven out of the generator flask by a slight stream of nitrogen,
conducted through 2 N sulfuric acid (to remove ammonia) and a
tube containing phosphorous pentoxide (to remove water) and
led into the solution of CA by a pvc tube. After the introduction
of cyclopropene had taken place for 10 min, the mixture was
irradiated by utilising a pyrex immersion well containing a
Hanovia mercury lamp (medium pressure, 450 W), which was
surrounded by a glass filter supposed to prevent the passage of
light of λ < 400 nm, at −30 °C. Initially, cyclopropene caused a
clouding (probably a precipitate of CA), but the mixture became
Compound 2, colourless crystals, mp 172–174 °C (Found: C,
47.7; H, 3.3. C14H12Cl4O2 requires C, 47.5; H, 3.4%); νmax
(KBr)/cm−1 2985w, 2963w, 2942w, 2930w, 1732s, 1726s,
1182m, 1099m, 1071w, 1036w, 1019w, 751w and 669w; δH
(250 MHz; CDCl3) 1.79 (2 H, m), 2.37–2.77 (6 H, m), 3.38
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 4400–4406 | 4405

View Article Online
3 (a) M. Christl, M. Braun, O. Deeg and S. Wolff, Eur. J. Org. Chem.,
2011, 968–982; (b) M. Braun, M. Christl, E.-M. Peters and K. Peters, J.
Chem. Soc., Perkin Trans. 1, 1999, 2813–2820; (c) M. Braun, M. Christl,
O. Deeg, M. Rudolph, E.-M. Peters and K. Peters, Eur. J. Org. Chem.,
1999, 2093–2101.
4 J.-H. Xu, Y.-L. Song, Z.-G. Zhang, L.-C. Wang and J.-W. Xu, Tetrahe-
dron, 1994, 50, 1199–1210.
5 S. W. Benson, Thermochemical kinetics, Wiley, New York, 2nd edn,
1976, p. 273.
clear again until the end of the cyclopropene addition (2 h).
After CA had been completely consumed (6 h of irradiation), the
mixture was quickly (15 min) concentrated to a volume of about
10 cm3 at 35 °C (bath)/15 mmHg, then immediately treated with
anhydrous methanol (150 cm3) and kept at 20 °C for 24 h. Evap-
oration of the methanol in vacuo and flash chromatography
(SiO2, pentane–ethyl acetate 9 : 1) of the residue at −30 °C furn-
ished (2α,5β,5aβ,6aβ,7β)-5,7,8-trichloro-3,5,5a,6,6a,7-hexahy-
dro-2-methoxy-5,7-methano-2H-cyclopropa[4,5]cyclohepta[1,2-
b]pyran-9-one 10 (375 mg, 36%) as an oil, which gave colour-
less crystals, mp 148–150 °C, after treatment with a small quan-
tity of ethyl acetate (Found: C, 48.2; H, 3.3. C13H11Cl3O3
requires C, 48.55; H, 3.45%); νmax (KBr)/cm−1 3085w, 3072w,
3002w, 2960m, 2940m, 2910w, 2900w, 2845w, 1788s, 1584m,
1451m, 1442m, 1420m, 1392m, 1373m, 1352m, 1338m,
1232m, 1222m, 1182m, 1172m, 1150m, 1142m, 1128m, 1113s,
1086m, 1068s, 1060s, 1022s, 1010s, 984s, 955m, 940s, 915s,
878s, 855m, 839m, 822s, 818s, 792w, 762w, 751m and 737m;
δH (200 MHz; CDCl3) 0.78 (1 H, dt, J 7.9 and 3.9, 6-Hα), 1.14
(1 H, dt, J 7.9 and 7.1, 6-Hβ), 2.04 (1 H, ddd, J 7.6, 7.1 and 3.8)
and 2.63 (1 H, ddd, J 7.6, 7.1 and 4.0) (5a-H and 6a-H), 2.48 (1
H, ddd, J 18.6, 5.6 and 2.0) and 2.68 (1 H, ddd, J 18.6, 4.3 and
3.4) (3-Hα and 3-Hβ), 3.42 (3 H, s, CH3), 5.16 (1 H, ddd, J 4.3,
2.0 and 0.6, 2-H) and 6.14 (1 H, ddd, J 5.6, 3.4 and 0.6, 4-H);
δC (50 MHz; CDCl3) 11.3 (C-6), 27.1 and 32.0 (C-5a and C-6a),
30.0 (C-3), 56.0 (CH3), 75.2 and 76.6 (C-5 and C-7), 98.6 (C-2),
115.9 (C-8), 118.5 (C-4), 130.7 (C-4a), 143.3 (C-8a) and 192.2
(C-9); m/z (EI, 70 eV) 326, 324, 322, 320 (M+, 1, 10, 30, 31%),
285 (10), 255 (10), 253 (15), 249 (15), 227 (11), 225 (14), 199
(12), 162 (11), 115 (14), 71 (79), 58 (100) and 45 (17).
6 J.-H. Xu, L.-C. Wang, J.-W. Xu, B.-Z. Yan and H.-C. Yuan, J. Chem.
Soc., Perkin Trans. 1, 1994, 571–577.
7 Empirical correlations between oxidation and ionisation potentials have
been established by: L. L. Miller, G. D. Nordblom and E. A. Mayeda, J.
Org. Chem., 1972, 37, 916–918, as well as P. G. Gassman and
R. Yamaguchi, J. Am. Chem. Soc., 1979, 101, 1308–1310.
8 P. Bischof and E. Heilbronner, Helv. Chim. Acta, 1970, 53, 1677–1682.
9 K. B. Wiberg, G. B. Ellison, J. J. Wendoloski, C. R. Brundle and N.
A. Kuebler, J. Am. Chem. Soc., 1976, 98, 7179–7182.
10 K. B. Wiberg, Angew. Chem., 1986, 98, 312–322; K. B. Wiberg, Angew.
Chem., Int. Ed. Engl., 1986, 25, 312–322.
11 (a) J. D. Roberts and R. H. Mazur, J. Am. Chem. Soc., 1951, 73, 2509–
2520; (b) K. L. Servis and J. D. Roberts, Tetrahedron Lett., 1967, 1369–
1372.
12 T. W. Bentley, S. J. Norman, R. Kemmer and M. Christl, Liebigs Ann.,
1995, 599–608.
13 F. Thalhammer, U. Wallfahrer and J. Sauer, Tetrahedron Lett., 1990, 31,
6851–6854.
14 T. W. Bentley, S. J. Norman, E. Gerstner, R. Kemmer and M. Christl,
Chem. Ber., 1993, 126, 1749–1757.
15 R. L. Cook and T. B. Malloy, Jr., J. Am. Chem. Soc., 1974, 96, 1703–
1707.
16 M. Christl, Chem. Ber., 1975, 108, 2781–2791.
17 M. S. Baird, D. G. Lindsay and C. B. Reese, J. Chem. Soc. C, 1969,
1173–1178.
18 K. Peters, E.-M. Peters, M. Braun and M. Christl, Z. Kristallogr. – New
Cryst. Struct., 2000, 215, 47–48.
19 T. Ichino, S. M. Villano, A. J. Gianola, D. J. Goebbert, L. Velarde,
A. Sanov, S. J. Blanksby, X. Zhou, D. A. Hrovat, W. T. Borden and W.
C. Lineberger, Angew. Chem., 2009, 121, 8661–8663; T. Ichino, S.
M. Villano, A. J. Gianola, D. J. Goebbert, L. Velarde, A. Sanov, S.
J. Blanksby, X. Zhou, D. A. Hrovat, W. T. Borden and W. C. Lineberger,
Angew. Chem., Int. Ed., 2009, 48, 8509–8511.
The 1H NMR spectrum of the crude product prior to methano-
lysis indicated the presence of a complex mixture, whose major
component, apart from chlorobenzene, was most likely the
α-chloro ether 14; δH (250 MHz; CDCl3) 0.84 (1 H, dt, J 7.9
and 3.9) and 1.21 (1 H, dt, J 7.9 and 7.1) (6-H2), 2.16 (1 H, ddd,
J 7.7, 7.1 and 3.7) and 2.69 (1 H, ddd, 7.7, 7.1 and 4.1) (5a-H
and 6a-H), 6.24 (1 H, ddd, J 6.6, 2.6 and 1.0) and 6.39 (1 H, dt,
J 4.3 and 1.2) (2-H and 4-H); because of several multiplets in
the pertinent region, the signals of the 3-CH2 group could not be
unambiguously localised, but the most likely chemical shifts are
δ 2.90 and 3.00.
20 H. F. Bettinger, Angew. Chem., 2010, 122, 680–682; H. F. Bettinger,
Angew. Chem., Int. Ed., 2010, 49, 670–671.
21 H. M. R. Hoffmann, Angew. Chem., 1984, 96, 29–48;
H. M. R. Hoffmann, Angew. Chem., Int. Ed. Engl., 1984, 23, 1–19.
22 A. R. Matlin, P. M. Lahti, D. Appella, A. Straumanis, S. Lin, H. Patel,
K. Jin, K. P. Schrieber, J. Pauls and P. Raulerson, J. Am. Chem. Soc.,
1999, 121, 2164–2173.
23 C. J. Cramer and S. E. Barrows, J. Phys. Org. Chem., 2000, 13, 176–186.
24 A. R. Matlin, in CRC Handbook of Organic Photochemistry and Photo-
biology, ed. W. M. Horspool and F. Lenci, CRC Press, Boca Raton, 2nd
edn, 2004, ch. 81.
25 (a) G. L. Closs and K. D. Krantz, J. Org. Chem., 1966, 31, 638;
(b) P. Binger, P. Wedemann and U. H. Brinker, Org. Synth., 2000, 77,
254–259; Org. Synth., 2004, Coll. Vol. 10, 231–234.
Notes and references
26 (a) R. Köster, S. Arora and P. Binger, Liebigs Ann. Chem., 1973, 1219–
1235; (b) P. Binger, Synthesis, 1974, 190–192; (c) B. Halton and M.
G. Banwell, in The Chemistry of the Cyclopropyl Group, ed. S. Patai and
Z. Rappoport, Wiley, Chichester, 1987, part 2, ch. 21.
27 Earlier investigations on this subject are quoted in ref. 3a.
28 Cyclobutene was prepared according to A. C. Cope, A. C. Haven, Jr., F.
L. Ramp and E. R. Trumbull, J. Am. Chem. Soc., 1952, 74, 4867–4871.
29 P. H. M. Budzelaar, gNMR, version 4.0, Cherwell Scientific Publishing
Ltd., Oxford, 1995–1997.
1 (a) K. Maruyama and A. Osuka, in The Chemistry of Quinonoid Com-
pounds, ed. S. Patai and Z. Rappoport, Wiley, New York, 1988, ch. 13,
vol. 2, part 1; (b) D. Creed, in CRC Handbook of Organic Photochemis-
try and Photobiology, ed. W. M. Horspool and P.-S. Song, CRC Press,
Boca Raton, 1995, pp. 737–747; (c) A. Gilbert, in CRC Handbook of
Organic Photochemistry and Photobiology, ed. W. M. Horspool and F.
Lenci, CRC Press, Boca Raton, 2nd edn, 2004, ch. 87.
2 S. M. Hubig, T. M. Bockman and J. K. Kochi, J. Am. Chem. Soc., 1997,
119, 2926–2935.
4406 | Org. Biomol. Chem., 2012, 10, 4400–4406
This journal is © The Royal Society of Chemistry 2012