Surprising exocyclic regioselectivity in electrophilic additions to alkylidenecyclobutenes

Article abstract of DOI:10.1016/S0040-4020(02)00267-3

The addition of peracid, bromine, TCNE and singlet oxygen to alkylidenecyclobutenes shows a strong regioselectivity - in some cases regiospecificity - for the exocyclic double bond. Depending on the reagent, the endocyclic double bond will only undergo reaction upon the addition of a second equivalent of reagent. In the case of the first three reagents, this exocyclic regioselectivity is rationalized by invoking the formation of the more stabilized carbocation intermediate or transition state. Regarding ¹O₂, where polar mechanisms are rare, we attribute this exocyclic regioselectivity to the improper alignment and, hence, lack of reactivity of the endocyclic allylic hydrogen.

Full text of DOI:10.1016/S0040-4020(02)00267-3

TETRAHEDRON
Pergamon
Tetrahedron ꢀ8 52002) 3199±320ꢀ
Surprising exocyclic regioselectivity in electrophilic additions
to alkylidenecyclobutenes
p
Hillel Pizem, Ofer Sharon and Aryeh A. Frimer
Department of Chemistry, Bar-Ilan University, Ramat Gan 52900, Israel
Received 2 December 2001; revised 31 January 2002; accepted 28 February 2002
AbstractÐThe addition of peracid, bromine, TCNE and singlet oxygen to alkylidenecyclobutenes shows a strong regioselectivityÐin some
cases regiospeci®cityÐfor the exocyclic double bond. Depending on the reagent, the endocyclic double bond will only undergo reaction
upon the addition of a second equivalent of reagent. In the case of the ®rst three reagents, this exocyclic regioselectivity is rationalized by
invoking the formation of the more stabilized carbocation intermediate or transition state. Regarding O , where polar mechanisms are rare,
1
2
we attribute this exocyclic regioselectivity to the improper alignment and, hence, lack of reactivity of the endocyclic allylic hydrogen.
q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
addition with TCNE at the exocyclic C ±C double bond
0
1 1
to give spiroheptane 6 as the sole product 5Eq. 52)); sur-
prisingly, the endocyclic double bond proved unreactive.
We decided to explore the generality of this observation
with two other electrophilic additions to this system, peracid
epoxidation and bromination.
Previous research in our laboratory has explored the effect
of ring-strain on the mode, rate and direction of singlet
oxygen attack. In particular, we have focused on various
1
small ring ole®n systems in which ring-strain decreases or
1
h
develops as we proceed to product. Frimer and Weiss
studied the singlet oxygenation of alkylidenecyclobutenes
Eq. 51)) and observed exclusive formation of ene reaction
5
product 2, resulting from oxygen attack at the exocyclic
double bond and abstraction of H . There was no evidence
a
of oxygen attack at the endocyclic double bond and abstrac-
tion of H , which would have yielded cyclobutene 3.
ꢀ
2†
b
ꢀ
1†
2. Results and discussion
1
f±h
2.1. Peracid epoxidation
Frimer and Weiss
to do with the spacial alignment of H , which is displaced
argued that the absence of 3 may have
b
2.1.1. Monoepoxidation. When isopropylidenecyclobutene
4 was reacted with 1 equiv. of m-CPBA in CH Cl or
ca. 368 from the perpendicular, far from the pseudo-axial
orientation strongly preferred by singlet oxygen. However,
an alternate rationale is also possible: it is not the allylic
hydrogen H of 1 that is unreactive, but rather the C ±C
3
endocyclic double bond. Indeed, we have recently reported
that isopropylidenecylobutene 4 undergoes [212] cyclo-
2
2
2
1
CH CN, the H NMR spectrum of the reaction mixture
3
following aqueous work-up showed the presence of a sole
product which contained the typical cyclobutene ole®nic
hydrogen absorption at 6.67 ppm. This suggested that
indeed the endocyclic double bond had remained intact.
b
2
3
1
3
However, the C NMR spectrum of this sample lacked
absorptions corresponding to epoxide carbons 5ca. 60±
6ꢀ ppm), revealing instead C±O absorptions at 74 and
89 ppm. In addition, other peaks strongly suggested the
Keywords: addition reactions; cyclobutenes; regioselection; oxygen,
singlet.
Corresponding author. Tel.: 1972-3-ꢀ318610; fax: 1972-3-ꢀ3ꢀ12ꢀ0;
e-mail: frimea@mail.biu.ac.il
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020502)00267-3

3200
H. Pizem et al. / Tetrahedron 58 '2002) 3199±3205
incorporation of a m-chlorobenzoic acid moiety into the
observed product. Silica gel chromatography of the reaction
mixture gave a 30% yield of a new isomeric product which
again contained a cyclobutene ole®nic hydrogen absorption
at 6.39 ppm, but had C±O absorptions at 7ꢀ and 80 ppm.
Based on this and other spectral data 5vide infra), we suggest
that the initially formed labile epoxide 8 5see Eq. 53)) under-
goes acid-catalyzed opening and nucleophilic attack by
m-chlorobenzoic acid at C-1, yielding butenol 10. The latter
undergoes silica mediated intramolecular trans-esteri®ca-
tion to the sterically less congested isomer 11.
as the sole product. Such a process is in fact well documen-
ted in the acid catalyzed solvolysis of allylic epoxides.
ꢀ
Nevertheless, we might have expected to see, in addition
to 10, products of S 1 attack. Interestingly, when the epoxi-
0
dation was repeated in CH OH, two new products were
N
3
obtained in a 2:1 ratio and identi®ed as methoxycyclo-
butenes 15 and 16 5Eq. 5ꢀ)).
ꢀꢀ†
53)
Long-range heteroCOSY experiments con®rmed these
assignments: in 15, cross-peaks were observed between
the ortho aromatic hydrogens and C-3, as well as between
the methyl hydrogens and C-1; in 16, cross-peaks were
found between the ring methylene hydrogens 5H-4) and
the ipso aromatic carbon, as well as between the methyl
and methoxy hydrogens and adjacent ring carbon C-1.
While the minor product 16 presumably results, as
Our rationale for the assignments of 10 and 11 is based on
1
h
the following considerations. Frimer and Weiss reported
that the m-CPBA epoxidation of the 3-butyl analog, cyclo-
butene 12, yields the corresponding epoxide 13. We have
reexamined their original spectral data and now believe the
authors erred; the isolated product is in fact diol 14a
0
before, from S 1 attack at C-1 of 9, the major product 15
N
results from S 1 attack at the less sterically hindered
N
0
5
Eq. 54)). The Dd between C-1 and C-1 is 17 ppm.
carbon C-3.
The solvent dependent modes of action observed may well
be linked to the protic nature of the media. In aprotic
solvents such as CH Cl and CH CN, the hydrogen bonding
5
4)
2
2
3
0
5pre-coordination) of the C-1 hydroxyl group with the
attacking carboxylic acid plays a dominant role in directing
the nucleophile to the C-1 position of 9 5Eq. 53)). Such
ꢀ
a±c
directing effects have been previously observed,
are expected to be attenuated in protic media, thus allowing
a competing S 1 approach.
N
but
4
Calculations indicate that replacing the butyl group in 14a
0
1
3
with a phenyl group 514b) has a negligible effect on the C
NMR chemical shifts of C-1 and C-1 . We also used these
0
calculations to explore the effect of esteri®cation of diols
2.1.2. Diepoxidation. When alkylidenecyclobutene 4 was
treated with 2 equiv. of m-CPBA in aprotic media, a new
product was formed whose mass corresponded to the
addition of two oxygen atoms and m-chlorobenzoic acid.
The same product was obtained when monoepoxidation
product 10 was treated with m-CPBA. After much spectral
analysis, this product was identi®ed as oxabicyclo[2.1.1]-
hexane 17, and a plausible mechanism of formation is
0
C-1 esteri®cation is predicted to increase the Dd between
1
4a and 14b at either the C-1 or C-1 hydroxyl groups. Thus,
0
0
C-1 and C-1 5by ca. 7 ppm), while C-1 esteri®cation would
decrease it 5by 8 ppm). These calculations con®rm that,
in the epoxidation of 4, the initially formed ester
possessing the larger Dd 51ꢀ ppm) is in all probability the
1
-benzoyl analog, butenol 10; the rearrangement product
0
²
shown in Eq. 56).
with the smaller Dd 5ꢀ ppm) is presumably the 1 -benzoyl
isomer 11.
In the mechanism outlined in Eq. 53), considering the 38
structure of the C-1 carbon of epoxide 4, we have posited
the intermediacy of an allylic carbonium ion 9, which
²
To aid the reader in following the transformation of 4 into 17, we have
kept the numbering of the carbons in these two compounds the same in
Eqs. 56) and 57). The correct numbering is shown in Scheme 2 and used in
Section 4.
undergoes S 1 attack by m-chlorobenzoic acid to yield 10
N

H. Pizem et al. / Tetrahedron 58 '2002) 3199±3205
3201
gens geminal to hydroxyl groups. The corresponding effect
observed for benzoyloxy is 3.3 ppm.
7
In summary, we see with m-CPBA a sizable difference in
reactivity rates of the endocyclic and exocyclic bonds of
alkylidenecyclobutene 4. In particular, no diepoxidation
product is observed until the exocyclic bond has fully
reacted.
56)
2.2. Bromination
Despite the fact that bromine is reported to react extremely
rapidly with styrene, 1 equiv. of bromine added exclusively
8
to the exocyclic double bond of alkylidenecyclobutene 4,
yielding dibromide 19 5Eq. 58)). Only the second equivalent
reacts at the endocyclic styrenyl double bond of 4
generating tetrabromide 20.
The NMR spectral data for 17, despite its relative simplicity,
is unique. The H NMR spectrum of the product revealed
1
three vicinally coupled aliphatic hydrogens at 2.87 5t,
C
1
3
Jˆ8 Hz), 3.ꢀ0 5d, Jˆ8 Hz) and ꢀ.035d, Jˆ8 Hz). The
NMR data further indicated the presence of a methylene and
a methyne at 41.12 and 83.27, respectively. The proposal of
structure 17 as drawn in Eq. 56) requires two atypical
splitting constants: 51) a relatively small geminal coupling
ꢀ
8†
5
8 rather than 12±1ꢀ Hz) for the methylene; 52) a strong
W-coupling 58 rather than 1 Hz) between one of these
methylene hydrogens and the methyne hydrogen across
the ring 5see Eq. 57)). Importantly, while such values are
generally speaking unusual, they are well precedented for
bicyclo[2.1.1]hexane systems. NOESY cross-peaks, too,
con®rm the structure proposed for 17. As further outlined
Attempts to purify 19 via silica gel column chromatography
yielded the known cyclopentenone 22 5Eq. 59)), pre-
9
6
sumably via silica mediated hydrolysis to a b-hydroxy-
bromide, cyclobutenol 21, which ring opens to the
observed cyclopentenone. Ghosez and coworkers simi-
in Eq. 57), the W-coupled hydrogens H4
0
and H show NOE
2
10
interaction with different C-ꢀ methyl groups, only possible
in such a bicyclo structure. This NOE interaction in turn
places the C-2 hydroxyl group endo in the bicyclo-system
larly report that when b-hydroxyselenide 23 is reacted
with base, the corresponding cyclopentenone is obtained
5Eq. 59)).
5
see Eq. 57)).
ꢀ
9†
ꢀ
7†
Thus, once again, we see a sizable difference in reactivity
rates between the endocyclic and exocyclic bonds of alkyl-
idenecyclobutene 4.
2.3. Reaction with TCNE
The last issue that requires elucidation is the location of the
benzoate group. We have placed it at C-1 rather than C-2 for
the following reasons: 51) The ester group was situated at
C-1 in both the starting material and the intermediate mono-
epoxidation product 10, precursors to 17. The C±O bonds at
C-2 and C-1 are nearly perpendicular to one another and
trans-esteri®cation is, hence, highly unlikely. 52) No long
range hetero-COSY interaction was observed between H400
The preference for electrophilic reaction at the exocyclic
bond was observed once again in the case of TCNE addi-
tion. As noted above 5Eq. 52)), 1-isopropylidene-3-phenyl-
2-cyclobutene 4 524a in Eq. 510)), yields spiro adduct 6
3
525a) exclusively. Similarly, we have found that the
3-butyl analog 12 524b) reacts with TCNE preferentially
at the exocyclic double bond, yielding adducts 25 and 26
1
1
and the benzoate carbonyl. 53) The H4
appears at 2.87 ppm, while the corresponding H absorption
0
methylene resonance
in a 4:1 ratio. There are also several previous reports that
variously ring-substituted methylenecyclobutenes 5e.g. 27
in Eq. 511)) react with TCNE via a facile 212 cycloaddition
exclusively at the exocyclic ole®nic linkage.
2
is down®eld at ꢀ.03ppm. This down®eld shift of 2.2 ppm is
slightly below the 2.ꢀ ppm down®eld shift typical for hydro-

3202
H. Pizem et al. / Tetrahedron 58 '2002) 3199±3205
ꢀ
10†
ꢀ
11†
Scheme 1. Calculated electron density on selected alkylidenecyclobutenes.
1
1c
In this light, the 40 year old report of Williams regarding
the reaction of 3-methyl-1-isopropylidenecyclobutene 524c)
with TCNE is highly surprising. In particular, he reports that
TCNE adds primarily at the internal cyclobutene double
bond, yielding 25 and 26 in a 1:4 ratio 5Eq. 510)). In com-
paring these latter results with the exocyclic preference of
is the stability of the zwitterionic intermediate. As shown in
Eq. 512), addition of TCNE to the exocyclic bond of an
alkylidenecyclobutene yields zwitterion 29, while reaction
at the endocyclic ole®nic linkage yields 30. The substitution
pattern in the former involves an allylic carbonium ion
which in the case of 24a 54) is benzylic at one end and 38
at the other; while in the case of 24b 512) and 27, it is 38 at
both ends. By contrast, 30 is an allylic carbonium ion which
is 28 and 38 in the case of 24, and 28 and 18 in the case of 27.
The stability difference between the various carbocations is
well known to be substantial, based on heterolytic C±H
dissociation energies in the gas phase. Thus 38, benzylic
and 28 are more stable than 18 carbocations by 4ꢀ, 39 and
1
1c
12
2
7, Williams and others had suggested that the steric
bulk of the gem-dimethyl groups in 24c inhibits approach to
the exocyclic bond. This rationale, however, is inconsistent
with our recent results from both 24a 54) and 24b 512).
Assuming the correctness of Williams report, we have not
found an alternative explanation.
2
.4. The question of mechanism
1
6
27 kcal/mol, respectively.
The observed exocyclic regioselectivity is somewhat
1
3
surprising. Doering and coworkers studied the heats of
hydrogenation of various cycloole®ns and report that
exocyclic unsaturation in four-membered ring compounds
is intrinsically slightly more stable 5by ca. 0.6 kcal/mol)
than endo unsaturation. In the speci®c case of ole®ns 24,
this difference should be increased by another 0.3kcal/
1
4
mol because the exocyclic double bond is tetrasubstituted
while the endocyclic unsaturation is trisubstituted. Thus the
endocyclic double bond, which is about 1 kcal/mol less
stable than the exocyclic one, should be more reactive.
5
12)
We also explored the possibility that these electrophilic
processes are kinetically controlled by the electron richness
of the respective double bonds. To this end, we calculated
p
the atomic charges using a 6-31G basis set and Mulliken
population analysis of the ole®nic carbons in various alkyl-
idenecyclobutenes 5see Scheme 1). The results would
seem to indicate that for compounds 24 5in contradistinction
to 27) it is speci®cally the endocyclic double bond which is
the more electron rich. Analysis of the orbital coef®cients
yields similar results. We are perforce lead to the con-
clusion that the controlling factors are not ground-state
considerations.
Assuming that the exocyclic regioselectivity of the addition
of TCNE to alkylidenecyclobutenes is driven by the stability
of the intermediate carbocation, a similar mechanism can
also be invoked to explain comparable results for the
electrophilic bromination of this conjugated diene system.
1
h
17
Indeed, Wilkins and Regulski report that in a similar
conjugated system, styrenes, bromination via a more open
carbonium ion intervenes. Finally, turning to the reaction of
m-CPBA with alkylidenecyclobutenes, the literature records
that peracid epoxidations of ole®ns generally proceed via a
one-step process which presumably lack carbocation
In the case of TCNE additions, the exocyclic regio-
selectivity is readily rationalized considering that the
mechanism for [212]cycloadditions of TCNE with ole®ns
1
8
character in the transition state. Nevertheless, a secondary
kinetic isotope effect study by Hanzlik and Shearer does
indicate that the transition state for the epoxidation of
1
ꢀ
19
are now well established to be non-concerted, ionic
processes. It is likely, therefore, that the controlling factor

H. Pizem et al. / Tetrahedron 58 '2002) 3199±3205
3203
Scheme 2. Numbering of the carbons used in the NMR spectral data.
arylethylenes is unsymmetrical and polar, in which only one
C±O bond of the oxirane product is formed. We suggest that
a similar charge development is involved in our 1,3-diene
system as well.
5AC-200), and via long range hetero COSY and NOESY
experiments 5DMX-600) as needed. In all cases, TMS
served as the internal standard. The carbon numbering of
the various compounds used in the nomenclature and spec-
tral assignments is shown in the Scheme 2. EI and CI 5CH4)
mass spectra were run on a Finnigan-4021 GC/MS machine
5at 70 eV, unless otherwise indicated), except where exact
mass data is given. In the latter instance, the EI data reported
is based on the high resolution mass spectra 5HRMS),
performed on a VG-Fison AutoSpecE High Resolution
Spectrometer. Preparative thin layer chromatography
5PTLC) was carried out on Merck silica gel F2ꢀ4 precoated
plates and the products were extracted from the silica by
3. Conclusion
The addition of TCNE, bromine, peracid and singlet oxygen
to alkylidenecyclobutenes 24 shows a strong preference for
reaction at the exocyclic double bond. For the ®rst three
reagents, we have rationalized this selectivity based on the
formation of the more stabilized carbocation. Such polar
mechanisms, however, are rare in the O ene reactions of
1
stirring overnight in dichloromethane or 10% CH
CHCl . For column chromatography separation, Merck
OH in
3
2
ole®ns and dienes, and are observedÐif at allÐin systems
substituted with heteroatoms. We must perforce return to
3
2
silica gel 230±400 mesh was used and the products were
extracted from the silica by stirring overnight in a solution
of. The retention times given are for the analytical runs.
1
the aforementioned suggestion of Frimer and Weiss who,
h
as noted in the Introduction, attributed this exocyclic regio-
selectivity of O with 1 to the unfavorable alignment and,
1
3
1h
Cyclobutenes 4 524a) and 12 524b) were prepared as
previously described.
2
hence, lack of reactivity of the endocyclic allylic hydrogen
H 5Eq. 51)).
b
4.2. General epoxidation procedure
4. Experimental
m-CPBA 51 or 2 equiv.) was added in one portion to a
magnetically stirred solution of cyclobutene 4 524a;
4
.1. General
1
equiv.) in an appropriate solvent 5ca. 2ꢀ mL of solvent
NMR spectra were obtained on Bruker DMX-600, DPX-300
and AC-200 Fourier transform spectrometers. Assignments
were facilitated with DEPT 5DPX-300), by correlating
proton and carbon chemical shifts through analysis of
residual couplings in off-resonance decoupled spectra
per mmol of substrate). After 10 min, the solution was trans-
ferred to a separatory funnel, extracted consecutively with a
10% bisul®te solution, with a saturated bicarbonate solu-
tion, and with a saturated sodium chloride solution, dried
over magnesium sulfate, and ®nally rotary evaporated.

3204
H. Pizem et al. / Tetrahedron 58 '2002) 3199±3205
1
5MH 2H±CH or M2CH , 24.ꢀ7%), 201 5MH 2H O,
1
4
.2.1.
buten-1-yl m-chlorobenzoate ꢀ10); 1-ꢀ1-m-chlorobenzoyl-
oxy-1-methylethyl)-3-phenyl-2-cyclobuten-1-ol ꢀ11).
1-ꢀ1-Hydroxy-1-methylethyl)-3-phenyl-2-cyclo-
3
3
2
1
10.21%), 187 5MH 2MeOH, 100%), 186 5M 2MeOH),
1
1
171 5M 2MeOH±CH ). HRMS 5CI, CH ): MH , found
1
3
4
Cyclobutene 4 524a) in dichloromethane was epoxidized
with 1 equiv. of m-CPBA as described in Section 4.2, yield-
ing ester 10 as a white solid in a 7ꢀ% yield. The latter can be
further puri®ed on reverse phase preparative TLC plate
eluting with 30% water in acetonitrile. Silica gel chromato-
graphy of the reaction mixture yielded as a white solid
which proved to be a 7:3mixture of 10 and isomeric benzo-
ate 11. The spectral data for the latter was extracted from
that of the mixture. Compound 10: dH 5300 MHz, CDCl3)
219.1390. C H O requires 219.138ꢀ.
1
4
19
2
4.2.3. 4-ꢀm-Chlorobenzoyloxy)-3,3-dimethyl-1-phenyl-2-
oxa[2.1.1]bicyclohexane-5-ol ꢀ17). Cyclobutene 4 in
CH Cl was reacted with 2 equiv. of m-CPBA as described
2
2
in Section 4.2. After 2 days of reaction, workup afforded
oxabicyclo 17 in ꢀ4% yield as a white solid. The product
can be further puri®ed on a preparative TLC plate eluting
with 10% 2-methyl-2-butanol in hexane. Compound 17: dH
5600 MHz, CDCl ) 8.035s, 1H, H ), 7.9ꢀ 5d, Jˆ8 Hz, 1H,
8
.035s, 1H, H 11), 7.98 5d, Jˆ8 Hz, 1H, H ), 7.42 5d,
10
3
11
Jˆ8 Hz, 1H, H ), 7.39 5m, 6H, aryl1H ), 6.72 5s, 1H,
H ), 7.ꢀ9 5d, Jˆ8 Hz, 1H, H ), 7.40 5m, 6H, aryl1H ),
1
4
12
14
13
1ꢀ
H ), 3.29 and 3.03 5ABq, Jˆ13Hz, 2H, H ), 1.39 and
ꢀ.035d, Jˆ8 Hz, 1H, H ), 3.ꢀ0 5d, Jˆ8 Hz, 1H, H 00), 2.87 5t,
0
) 1.ꢀ1 and 1.49 5each s, each 3H, C and C
6 7 8
2
4
ꢀ
6
1.36 5each s, each 3H, C and C methyls); d 57ꢀ.ꢀ MHz,
CDCl ) d 166.28 5C ), 148.64 5C ), 134.ꢀ9 5C ), 133.14
Jˆ8 Hz, 1H, H
methyls); dC 51ꢀ0.9 MHz, CDCl ) d 16ꢀ.ꢀ35C ), 134.77
3 9
6
7
C
3
8
3
9
5
and 5C , C and C ), 128.ꢀ0 5meta), 127.935 para),
C ), 132.47 and 132.71 5C and ipso), 129.72, 129.84
5C ), 134.38 5C ), 133.80 5C ), 131.0ꢀ 5C ), 130.01
10 16 13 12
5C11 and C ), 128.30 5C and C ), 128.14 5C ), 126.67
1ꢀ 18 19 14
1
4
13
1
0
11
12
1
26.02 5C ), 12ꢀ.61 5ortho), 89.14 5C ), 73.68 5C ), 38.17
2
5C ), 84.31 5C ), 83.27 5C ), 82.74 5C ), 77.ꢀ35C ), 41.12
17 3 ꢀ 1 4
1
ꢀ
5
5
C ), 2ꢀ.71 and 2ꢀ.91 5C and C ); m/z 5DCI, CH ) 3 4ꢀ
2
5C ), 24.70 and 23.ꢀ4 5C and C ); m/z 5GC/MS, CI, CH )
6 7 8 4
6
7
4
1
1
MH 12, 1.3%), 343 5MH , 4%), 32ꢀ 5MH 2H O,
1
1
1
361 5MH 12, 3.04%), 3ꢀ9 5MH , 9.14%), 2035MH 2m-
1
2
1
1
1%), 284 5MH 2i-PrOH, 44%), 187 5MH 2chloro-
1
chlorobenzoic acid, 100%), 2035MH 2chlorobenzoic
2
benzoic acid); HRMS 5DCI, CH ): MH , found 343.1078.
1
1
acid±H O, ꢀ0.ꢀ1%); HRMS 5CI, CH ): MH , found
4
2
4
C H O Cl requires 343.1100. Compound 11: dH
2
3ꢀ9.1090. C H O requires 3ꢀ9.10ꢀ0.
14 19
0
20
3
2
5
300 MHz, CDCl ) 8.10 5s, 1H, H ) 8.09 5d, Jˆ8 Hz, 1H,
10 14 12
3
11
H ) 7.60 5d, Jˆ8 Hz, 1H, H ) 7.40 5m, 6H, aryl1H ) 6.39
4.2.4. 1-Bromo-1-ꢀ1-bromo-1-methylethyl)-3-phenyl-2-
cyclobutene ꢀ19); 5,5-dimethyl-3-phenyl-2-cyclopenten-
1-one ꢀ22). Bromine 50.ꢀ4 mmol) in CH Cl 51.1ꢀ mL of
5
s,1H, H ) 3.11 and 2.63 5ABq, Jˆ13Hz, 2H, H ) 1.48 and
4
2
1
CDCl ) 164.33 5C ), 79.ꢀ1 5C ), 74.82 5C ), 40.16 5C ),
.42 5each s, each 3H, C and C methyls); d 57ꢀ.ꢀ MHz,
6 7 C
2
2
0.468 M bromine solution) was added dropwise to a stirred
solution of cyclobutene 4 591.7 mg, 0.ꢀ3mmol) in CH Cl
2
3
8
1
ꢀ
2
2
2.29 and 21.4ꢀ 5C and C ).
7
6
2
520 mL). The solvent was removed in vacuo, and the residue
4
1
.2.2. 1-ꢀ1-Hydroxy-1-methylethyl)-3-methoxy-3-phenyl-
-cyclobutene ꢀ15); 1-ꢀ1-hydroxy-1-methylethyl)-1-
recrystallized from hexane, yielding pure dibromocyclo-
butene 19 as a white solid 5100 mg, 3.06 mmol, ꢀ8%
yield). Attempts to purify 19 via column chromatography
methoxy-3-phenyl-2-cyclobutene ꢀ16). Cyclobutene 4
9
5
5
280 mg, 1.64 mmol), dissolved in dichloromethane
2 mL) and methanol 540 mL), was epoxidized with
yielded the known cyclopentenone 22. Compound 19: dH
5300 MHz, CDCl ) 7.41±7.37 5m, ꢀH, aryl), 6.ꢀ2 5s, 1H,
3
1 equiv. of m-CPBA as outlined in the general epoxidation
procedure. Work-up yielded a mixture of 15 and 16
C ), 3.ꢀꢀ and 3.3ꢀ 5Abq, Jˆ14 Hz, 2H, C ), 2.02 and 2.01
2
4
5each s, each 3H, C and C ); d 57ꢀ.ꢀ MHz, CDCl ) 14ꢀ.87
C
6
7
3
5
5
290 mg, 1.33 mmol, 81%) as a white solid in a 2:1 ratio
by NMR). The NMR spectral data for each was readily
5C ), 132.3ꢀ 5ipso), 131.107 5C ), 129.42 5para), 128.ꢀ6
3 2
5meta), 12ꢀ.62 5ortho), 73.47 5C ), 69.4ꢀ 5C ), 4ꢀ.6ꢀ 5C ),
1
ꢀ
4
extracted from that of the mixture. The MS and HRMS
data for each component was obtained with the help of
GC/MS using a 0.32 mm£30 m DB-23 5J&W Scienti®c)
fused silica capillary column, temperature programmed
from 130 to 2008C. Compound 15: dH 5600 MHz, CDCl3)
31.4ꢀ and 31.3ꢀ 5C and C ); m/z 5EI) 332 5M14, 3.42%),
6 7
1
330 5M12, ꢀ.04%), 328 5M , 3.ꢀ2%), 2ꢀ1 5C H Br,
1
3
14
1
23.76%), 170 5C H , 100%); HRMS 5EI): M , found
1
3
14
1
327.9488. C H Br 5M ) requires 327.9462.
1
3
14
2
7
.41 5m, ꢀH, aromatic ring), 6.36 5s, 1H, H ), 3.33 5s, 3H,
2
4.2.5. 1,2,3-Tribromo-1-ꢀ1-bromo-1-methylethane)-3-
phenyl-2-cyclobutene ꢀ20). Bromine 50.66 mmol) in
H ), 2.90 and 2.71 5ABq, Jˆ12.ꢀ Hz, 2H, H ), 1.40 and 1.39
8
4
5each s, each 3H, C and C methyls); d 51ꢀ0.9 MHz,
CDCl ) 1ꢀ8.24 5C ) 141.91 5ipso), 128.87 5para), 128.37
CH Cl 51.4 mL of 0.468 M bromine solution) was added
2
6
7
C
2
dropwise to a stirred solution of cyclobutene 4 5ꢀ4.4 mg,
0.32 mmol) in CH Cl 520 mL). The solvent was removed
3
1
5
5
meta), 12ꢀ.47 5ortho), 126.84 5C ), 79.79 5C ), 69.ꢀ1
2
C ), ꢀ2.41 5C ), 41.67 5C ), 27.3ꢀ and 27.43 5C and C );
ꢀ
1
2
2
in vacuo. Recrystallization from hexane gave tetrabromo-
cyclobutene 20 580 mg, 0.16ꢀ mmol, ꢀ0% yield) as a white
solid. Compound 20: dH 5300 MHz, CDCl ) 7.712±7.68ꢀ
3
8
4
6
7
1
1
m/z 5GC/MS, CI, CH ) 219 5MH , 14.ꢀ7%), 218 5M ,
4
1
.94%), 201 5MH 2H O, 100%), 187 5MH 2MeOH,
1
ꢀ
ꢀ
2
1
0.ꢀ1%), 169 5MH 2MeOH±H O), 160 5M 2i-PrOH);
1
5m, 2H, ortho), 7.414±7.371 5m, 3H, meta and para) 6.16
5d, Jˆ1 Hz, 1H, H ), 4.084 5dd, Jˆꢀ, 1 Hz, 2H, H ), 1.96ꢀ
and 1.941 5each s, each 3H, C and C methyls); dC
6 7
2
1
HRMS 5CI, CH ): MH , found 219.1400. C H O requires
4
14 19
2
2
4
2
ꢀ
1
19.138ꢀ. Compound 16: dH 5600 MHz, CDCl ) 7.41 5m,
3
H, aromatic ring), 6.33 5s, 1H, H ), 3.39 5s, 3H, H ), 2.845s,
2
H, H ), 1.29 5s, 3H, H ), 1.2ꢀ 5s, 3H, H ); d 51ꢀ0.9 MHz,
4
57ꢀ.ꢀ MHz, CDCl ) 139.00 5ipso), 129.00 5ortho), 128.6ꢀ
3
8
5para), 127.39 5meta), 73.ꢀ7 and 72.38 5C and C ), 62.97
3
6
7
C
1
CDCl ) 149.01 5C ) 133.16 5ipso), 128.87 5para), 128.48
3
5C ), ꢀ7.47 5C ), 49.36 5C ), 31.ꢀ4 and 30.31 5C and C );
2 ꢀ 4 6 7
3
5
5
meta), 12ꢀ.47 5ortho), 127.ꢀ0 5C ), 8ꢀ.36 5C ), 73.68
2
m/z 5EI): 494 5M18, 0.ꢀꢀ%), 492 5M16, 1.44%), 490
5M14, 1.90%), 488 5M12, 0.84%), 486 5M , 0.22%),
1
1
C ), ꢀ1.68 5C ), 33.73 5C ), 24.ꢀ0 and 22.71 5C and C );
ꢀ
8
4
6
7
1
m/z 5GC/MS, CI, CH ) 219 5MH , 11.46%), 203
4
411 5C H Br , 9ꢀ.29%), 329 5C H Br , ꢀ6.28%), 248
3
13 14
13 13
2

H. Pizem et al. / Tetrahedron 58 '2002) 3199±3205
320ꢀ
1
C H Br, 63.36%), 169 5C H , 100%), 1ꢀꢀ 5M 2Br ±
5
HBr±CH Br, ꢀ1.10%); HRMS 5EI): M 16, found
2. For reviews see: 5a) Frimer, A. A. Chem. Rev. 1979, 79, 3ꢀ9±
387. 5b) Frimer, A. A. In The Chemistry of Peroxides; Patai,
S., Ed.; Wiley: New York, 1983; pp 201±234. 5c) Frimer, A.
1
3
13
13 13
2
1
2
4
M 2H±Br , found 169.1010. C H requires 169.1017.
91.779ꢀ. C H Br requires 491.7769; HRMS 5EI):
13 14 4
1
A.; Stephenson, L. M. In Singlet O ÐVolume II: Reaction
2
4
13 13
Modes and ProductsÐPart I; Frimer, A. A., Ed.; Chemical
Rubber Company: Boca Raton, FL, 198ꢀ; pp 67±91.
5d) Frimer, A. A. In The Chemistry of EnonesÐPart 2;
Patai, S., Rappoport, Z., Eds.; Wiley: Chichester, 1989;
pp 781±921.
4
2
.2.6. 3,3-Dimethyl-6-butyl-spiro[3.3]-hept-5-ene-1,1,2,
-tetracarbonitrile ꢀ25b); 1-n-butyl-5-isopropylidenebi-
cyclo[2.2.0]hexa-2,2,3,3-tetracarbonitrile ꢀ26b). A ¯ame
dried two necked ¯ask equipped with a nitrogen inlet
adapter and glass stopper, was charged with a magnetically
stirred suspension of 1-isopropylidene-3-butyl-2-cyclo-
3. Frimer, A. A.; Pizem, H. Tetrahedron 1999, 55, 1217ꢀ±
12186.
3
13
4. Calculations of the C NMR data were made using the Chem-
butene 12 524b; 0.20 g, 1.33 mmol) in 30 ml dichloro-
1
3
methane. Tetracyanoethylene 5 50.242 g, 1.89 mmol) was
added in one portion yielding a blue solution, which turned
green after several minutes and yellow overnight. The
solvent was removed and the blue-gray residue 5297 mg)
was chromatographed over alumina eluting with 20%
ethyl acetate in petroleum ether yielding a pale orange
viscous liquid 5121 mg, 0.43ꢀ mmol, 33% yield,
Window35Softshell International)
developed by Dr Erno Pretsch 5ETH, Zurich).
C NMR module
ꢀ. See for example: 5a) Korach, M.; Nielsen, D. R.; Rideout,
W. H. J. Am. Chem. Soc. 1960, 82, 4328±4330. 5b) Sable,
H. Z.; Posternak, T. Helv. Chim. Acta. 1962, 45, 370±37ꢀ.
5c) Mayor, P. A.; Meakins, G. D. J. Chem. Soc. 1960,
2792±2800. 5d) Buchanan, J. G.; Sable, H. Z. Selective
Organic Transformations; Thyagarajan, B. S., Ed.; Wiley-
Interscience: New York, 1970; Vol. 2, p 1.
R ˆ0.603). Spectral analysis revealed it to be a 4:1 mixture
f
of tetranitriles 25b and 26b. Mixture of compounds 25b and
2
6b: nmax 5KBr) 2969, 2933, 2874, 2248, 2196, 171ꢀ, 163ꢀ,
6. 5a) Wiberg, K. B.; Lamoman, G. M.; Ciula, R. P.; Connor,
D. S.; Schertler, P.; Lavanish, P. Tetrahedron 1965, 21, 2749±
2769. 5b) Davies, H. G.; Roberts, S. M.; Wake®eld, B. J.;
Winders, J. A.; Williams, D. J. J. Chem. Soc., Chem. Commun.
1984, 640±640.
2
1
1
466, 1396, 1379, 1269, 1149, 8ꢀꢀ cm ; m/z 5CI, CH ) 279
4
1
1
5
MH , 77.0ꢀ%), 1735MH 2C H N , 62.63%), 1ꢀ1
6 6 2
1
1
5
MH 2C N , 28.21%), 1ꢀ0 5M 2C N , 100%); HRMS
6 4 6 4
1
5
CI, CH ): MH , found 279.1609. C H N requires
4 17 18 4
2
79.1610. Compound 25b: dH 5600 MHz, CDCl ) ꢀ.97
3
7. Gunter, H. NMR Spectroscopy; 2nd ed; Wiley: New York,
199ꢀ.
5
1H, t, Jˆ1.ꢀ Hz, H ), 2.81 and 2.78 52H, Abq of t,
ꢀ
Jˆ14.ꢀ, 1 Hz, H ), 2.17 52H, bt, Jˆ7.ꢀ Hz, H ), 1.61 53H,
8. Mare, D. Q. Rev. Chem. Soc. 1949, 3, 126±14ꢀ.
9. Iwasawa, N.; Matsuo, T.; Iwamoto, M.; Ikeno, T. J. Am.
Chem. Soc. 1998, 120, 3903±3914.
7
8
s, H ), 1.ꢀ4 53H, s, H ), 1.47 52H, quint, Jˆ7.ꢀ Hz, H ),
1
2
13
9
1
d
.36 52H, sex, Jˆ7.ꢀ Hz, H ), 0.935 3H , t, Jˆ7.ꢀ Hz, H );
51ꢀ0.1 MHz, CDCl ) 1ꢀ9.22 5C ), 124.11 5C ), 110.4ꢀ
C 3 6 ꢀ
1
0
11
10. Schmit, C.; Sahraoui-Taleb, E.; Dehasse-De Lombaert, C. G.;
Ghosez, L. Tetrahedron Lett. 1984, ꢀ043±ꢀ046.
5
CN), 110.04 5CN), 109.435CN), 109. 38 5CN), ꢀ7.12 5C 4),
ꢀ
2
0.32 5C ), 42.78 5C ), 41.435C ), 37.30 5C ), 30.33 5C ),
8
11. 5a) Blomquist, A. T.; Meinwald, Y. C. J. Am. Chem. Soc.
1957, 79, 4013±401ꢀ. 5b) Blomquist, A. T.; Meinwald, Y. C.
J. Am. Chem. Soc. 1959, 81, 667±672. 5c) Williams, J. K.
J. Am. Chem. Soc. 1959, 81, 4013±401ꢀ.
3
2
1
7
7.77 5C ), 24.00 and 23.92 5C and C ), 22.135C ), 13.62
13
9
12
10
5
C ); Compound 26b: d 5300 MHz, CDCl ) 3.86 5s, 1H,
11 H 3
H ), 3.3ꢀ and 3.0ꢀ 52H, Abq, Jˆ18 Hz, H ), 2.0ꢀ 52H, ddd,
4
6
Jˆ14, 11, ꢀ.ꢀ Hz, H
0
) 1.8ꢀ 5ddd, Jˆ14, 11, 6 Hz, H 00),
.73and 1.71 5each s, each 3H , C and C methyls), 1.43
8 9
12. Ciganek, E.; Linn, W. J.; Webster, O. W. In The Chemistry of
the Cyano Group; Rappoport, Z., Ed.; Wiley: Chichester,
1970; pp 423±669 at p 4ꢀꢀ.
1
0
10
1
5
d
2H, m, H ), 1.30 52H, m, H ), 0.96 53H, t, Jˆ7.ꢀ Hz, H );
1
2
11
13
51ꢀ0.1 MHz, CDCl ) 139.42 5C ), 117.97 5C ), 110.82
3
13. Turner, R. B.; Goebel, P.; Mallon, B. J.; von E. Doering, W.;
Colburn, Jr, J. F.; Pomerantz, M. J. Am. Chem. Soc. 1968, 90,
431ꢀ±4322.
C
ꢀ
7
5
4
5
CN), 109.61 5CN), 109.435CN), 109.01 5CN), ꢀ7.62 5C 4),
8.39 5C ), 4ꢀ.10 and 44.3ꢀ 5C and C ), 38.ꢀꢀ 5C ), 33.ꢀ4
C ), 2ꢀ.00 5C ), 22.30 5C ), 19.ꢀ8 and 18.98 5C and C ),
1
2
3
6
1
0
11
12
8
9
14. Kistiakowsky, G. B.; Ruhoff, J. R.; Smith, H. A.; Vaughan,
W. E. J. Am. Chem. Soc. 1936, 58, 146±1ꢀ3.
1
3.ꢀ7 5C ).
13
1
1
1
1
ꢀ. See Fatiadi, A. J. Synthesis 1987, 749±789 and references
cited therein.
References
6. March, J. Advanced Organic Chemistry; 4th ed; Wiley-
Interscience: New York, 1992; p 171, Table ꢀ.1.
7. Wilkins, C. L.; Regulski, T. W. J. Am. Chem. Soc. 1972, 94,
1
. 5a) Frimer, A. A.; Roth, D.; Sprecher, M. Tetrahedron Lett.
977, 1927±1930. 5b) Frimer, A. A.; Farkash, T.; Sprecher, M.
1
6
016±6020.
J. Org. Chem. 1979, 44, 989±99ꢀ. 5c) Frimer, A. A.; Roth, D.
J. Org. Chem. 1979, 44, 3882±3887. 5d) Frimer, A. A.;
Antebi, A. J. Org. Chem. 1980, 45, 2334±2340. 5e) Frimer,
A. A. Isr. J. Chem. 1981, 21, 194±202. 5f) Frimer, A. A.
J. Photochem. 1984, 25, 211±226. 5g) Frimer, A. A.; Weiss,
J. J. Org. Chem. 1993, 58, 3660±3667. 5h) Frimer, A. A.;
Weiss, J.; Gottlieb, H. E.; Wolk, J. L. J. Org. Chem. 1994,
8. 5a) March, J. Advanced Organic Chemistry; 4th ed; Wiley-
Interscience: New York, 1992; p 826. 5b) Plesnicar, B. In
The Chemistry of Peroxides; Patai, S., Ed.; Wiley: New
York, 1983; pp ꢀ21±ꢀ84 at p ꢀ28ff.
1
9. Hanzlik, R. P.; Shearer, G. O. J. Am. Chem. Soc. 1975, 97,
ꢀ
231±ꢀ233.
5
9, 780±792.