Unsubstituted bicyclo[1.1.0]but-2-ylcarbinyl cations

Article abstract of DOI:10.1021/jo0519918

A synthesis for the unsubstituted bicyclo[1.1.0]but-2-ylmethanols (endo- and exo-9) from 1,3-butadiene has been developed. Solvolyses of their sulfonates 10 and 11 took entirely different courses, as the endo compound 10 gave rise exclusively to rearranged products such as cyclopent-3-en-1-ol (14), while the exo compound 11 underwent only the substitution of the tosylate group with complete retention of the exo-bicyclo[1.1.0]but-2-ylmethyl skeleton. Under solvolytic conditions, 10 reacted at very similar rates to the corresponding monocyclic substrate, that is, cyclopropylcarbinyl mesylate (19); in contrast, 11 reacted only three times as fast as n-butyl tosylate and about 1000-fold slower than 10. The nature of the bicyclo[1.1.0]but-2-ylcarbinyl cations has been probed by quantum chemical calculations. Whereas, the exo isomer (exo-18) corresponds to a local energy minimum, the endo isomer is only a transition state [endo-l8(TS)] for an automerization of the nonclassical cyclopent-3-en-1-yl cation (13) and converts into 13 by a Wagner-Meerwein rearrangement. The most favorable isomerization of exo-18 also leads to 13 but via a transition state resembling the 2-vinylcycloprop-1-yl cation [25(TS)]. On the introduction of methyl groups at positions 1 and 3 of exo-18, the cation is no longer an energy minimum and it becomes a transition state [27(TS)] for an automerization of the nonclassical 1,3-dimethylcyclopent-3-en-1-yl cation (28). The large effect of the methyl substitution rationalizes the puzzling results of the previous product and rate studies, which utilized various substituted derivatives of bicyclo[1.1.0]but-2-ylcarbinyl sulfonates as substrates.

Full text of DOI:10.1021/jo0519918

Unsubstituted Bicyclo[1.1.0]but-2-ylcarbinyl Cations
T. William Bentley,*,† Bernd Engels,*,‡ Thomas Hupp,‡ Elena Bogdan,‡,§ and
Manfred Christl*,‡
Institut fu¨r Organische Chemie, UniVersita¨t Wu¨rzburg, Am Hubland, D-97094 Wu¨rzburg, Germany, and
Department of Chemistry, UniVersity of Wales, Swansea, Singleton Park,
Swansea SA2 8PP, United Kingdom
T.W.BENTLEY@swansea.ac.uk; bernd@chemie.uni-wuerzburg.de; christl@chemie.uni-wuerzburg.de
ReceiVed September 23, 2005
A synthesis for the unsubstituted bicyclo[1.1.0]but-2-ylmethanols (endo- and exo-9) from 1,3-butadiene
has been developed. Solvolyses of their sulfonates 10 and 11 took entirely different courses, as the endo
compound 10 gave rise exclusively to rearranged products such as cyclopent-3-en-1-ol (14), while the
exo compound 11 underwent only the substitution of the tosylate group with complete retention of the
exo-bicyclo[1.1.0]but-2-ylmethyl skeleton. Under solvolytic conditions, 10 reacted at very similar rates
to the corresponding monocyclic substrate, that is, cyclopropylcarbinyl mesylate (19); in contrast, 11
reacted only three times as fast as n-butyl tosylate and about 1000-fold slower than 10. The nature of the
bicyclo[1.1.0]but-2-ylcarbinyl cations has been probed by quantum chemical calculations. Whereas, the
exo isomer (exo-18) corresponds to a local energy minimum, the endo isomer is only a transition state
[endo-18(TS)] for an automerization of the nonclassical cyclopent-3-en-1-yl cation (13) and converts
into 13 by a Wagner-Meerwein rearrangement. The most favorable isomerization of exo-18 also leads
to 13 but via a transition state resembling the 2-vinylcycloprop-1-yl cation [25(TS)]. On the introduction
of methyl groups at positions 1 and 3 of exo-18, the cation is no longer an energy minimum and it
becomes a transition state [27(TS)] for an automerization of the nonclassical 1,3-dimethylcyclopent-3-
en-1-yl cation (28). The large effect of the methyl substitution rationalizes the puzzling results of the
previous product and rate studies, which utilized various substituted derivatives of bicyclo[1.1.0]but-2-
ylcarbinyl sulfonates as substrates.
Introduction
hex-3-yl sulfonates2 and tricyclo[4.1.0.02,7]hept-3-yl, as well as
-hept-4-en-3-yl esters,3 we studied the solvolysis of monosub-
stituted bicyclo[1.1.0]but-2-ylcarbinyl esters [dimesylates en-
do,endo- and exo,exo-4 (Scheme 2)]4 and found substantial
differences on comparison with the results for the diastereo-
isomers of 2. As expected, endo,endo-4 gave rise exclusively
to rearranged solvolysis products such as 5, but no rearrange-
ment occurred in the case of exo,exo-4, as illustrated by the
formation of the ethyl ether 6. Furthermore, endo,endo-4
underwent hydrolysis in 40% (v/v) acetone/water at 25 °C about
eight times as fast as exo,exo-4, that is, the exo/endo rate was
<1, despite the deactivating effect on the endo substrate of the
second CH2OMs group in close proximity.4
In 1970, Breslow et al.1 published a communication entitled
“Bicyclo[1.1.0]butyl-2-carbinyl Cations” and reported rate
constants of the solvolysis of the di- and trisubstituted bicyclo-
[1.1.0]but-2-ylcarbinyl tosylates 1 and 2 (Scheme 1). Whereas
no product structures were given in the case of 1, the
cyclopentenol 3 was described as the major compound formed
from both endo-2 and exo-2 on hydrolysis in 80% (v/v) dioxane/
water with an exo/endo rate ratio of 2 (the rate constants at 25
°C being 2.71 × 10-4 and 5.63 × 10-4 s-1, respectively).
After having investigated a number of endo,endo-bridged
bicyclo[1.1.0]but-2-ylcarbinyl esters, that is, tricyclo[3.1.0.02,6]-
† University of Wales.
‡ Universita¨t Wu¨rzburg.
§ On leave from the Faculty of Chemistry and Chemical Engineering, “Babes-
Bolyai” University, Cluj-Napoca, Romania.
(1) Breslow, R.; Bozimo, H.; Wolf, P. Tetrahedron Lett. 1970, 2395-
2397.
(2) Bentley, T. W.; Norman, S. J.; Gerstner, E.; Kemmer, R.; Christl,
M. Chem. Ber. 1993, 126, 1749-1757.
(3) Bentley, T. W.; Llewellyn, G.; Norman, S. J.; Kemmer, R.; Kunz,
U.; Christl, M. Liebigs Ann./Recl. 1997, 229-244.
10.1021/jo0519918 CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/06/2006
1018
J. Org. Chem. 2006, 71, 1018-1026

Unsubstituted Bicyclo[1.1.0]butyl-2-carbinyl Cations
SCHEME 1. Bicyclo[1.1.0]butyl-2-carbinyl Tosylates
Solvolyzed by Breslow et al.1
SCHEME 3. Synthesis of Tricyclo[4.1.0.02,7]heptan-3-ol (7)
According to Tischer6
SCHEME 4. Syntheses of endo-Bicyclo[1.1.0]but-2-ylcarbinyl
Mesylate (10) and exo-Bicyclo[1.1.0]but-2-ylcarbinyl Tosylate
(11)
SCHEME 2. Ethanolysis of the
Bicyclo[1.1.0]butan-2,4-dimethanol Dimesylates endo,endo-4
and exo,exo-44
To understand the differences between system 2 and system
4 regarding the rates as well as the effect of substituents on the
mechanism of the solvolysis reactions, we now report the first
investigation of the parent systems, namely, the unsubstituted
diastereomeric bicyclo[1.1.0]but-2-ylcarbinyl sulfonates, by
experimental methods and the corresponding bicyclo[1.1.0]but-
2-ylcarbinyl cations by quantum chemical calculations.
Results and Discussion
irane, which is the epoxidation reagent of choice with regard
to the purity of the products.8 The resulting 1:1 mixture of the
epoxides 8 was separated by chromatography, and the assign-
ment of their configuration was deduced from the outcome of
their reaction with 2 equivalents of n-butyllithium. It is assumed
that the anionic carbon atoms of the carbenoids formed from
the 8 isomers attack the epoxide subunits in an SN2-type manner,
which is why we take threo-8 as the precursor of endo-bicyclo-
[1.1.0]butane-2-methanol (endo-9) and erythro-8 as the precursor
of exo-9. Both alcohols 9 were obtained as colorless liquids,
from which an impurity of 1-butanol could not be removed
completely as a result of the sensitivity of the compounds. The
presence of the bicyclo[1.1.0]butane moieties in the products
emerging from 8 was derived from typical NMR data such as
δ ) 1.7 and -0.3 for C1,3 as well as JC1,H1 ) JC3,H3 ) 201.3
and 203.7 Hz for endo-9 and exo-9, respectively. The stereo-
chemical assignment is unambiguous on the basis of the
1. Synthesis of endo-Bicyclo[1.1.0]but-2-ylcarbinyl Mes-
ylate (10) and exo-Bicyclo[1.1.0]but-2-ylcarbinyl Tosylate
(11). endo- (endo-9) and exo-Bicyclo[1.1.0]butane-2-methanol
(exo-9). On the basis of the observations of Coates et al.,5 who
obtained homocyclopropylcarbinols by reductive cyclization of
bromocyclopropyl epoxides, Tischer6 developed the most ef-
ficient one among the syntheses of tricyclo[4.1.0.02,7]heptan-
3-ol (7),3 which is depicted in Scheme 3. We applied this
reaction sequence to the preparation of both bicyclo[1.1.0]-
butane-2-methanols 9 (Scheme 4). Accordingly, the known 1,1-
dibromo-2-vinylcyclopropane7 was treated with dimethyldiox-
(4) Bentley, T. W.; Llewellyn, G.; Kottke, T.; Stalke, D.; Cohrs, C.;
Herberth, E.; Kunz, U.; Christl, M. Eur. J. Org. Chem. 2001, 1279-1292.
(5) (a) Last, L. A.; Fretz, E. R.; Coates, R. M. J. Org. Chem. 1982, 47,
3211-3219. (b) Coates, R. M.; Last, L. A. J. Am. Chem. Soc. 1983, 105,
7322-7326.
(6) Tischer, W. Dissertation, Universita¨t Mu¨nchen, Germany, 1991.
(7) (a) Skell, P. S.; Garner, A. Y. J. Am. Chem. Soc. 1956, 78, 5430-
5433. (b) Huwyler, R.; Al-Dulayymi, A.; Neuenschwander, M. HelV. Chim.
Acta 1999, 82, 2336-2347.
(8) Adam, W.; Saha-Mo¨ller, C. R.; Zhao, C.-G. Org. React. 2002, 61,
219-516.
J. Org. Chem, Vol. 71, No. 3, 2006 1019

Bentley et al.
SCHEME 5. Solvolysis Products of
SCHEME 6. Solvolysis Products of
endo-Bicyclo[1.1.0]but-2-ylcarbinyl Mesylate (10)
exo-Bicyclo[1.1.0]but-2-ylcarbinyl Tosylate (11)
Scheme 5 illustrates the mechanism in which the reorganization
of the skeleton of 10 should proceed simultaneously with the
cleavage of the C-O bond. This view is supported by the
complete absence of the unrearranged products.
The solvolysis of 11 took an entirely different course, as only
products with a retained carbon atom skeleton were observed.
Thus, on dissolution in [D4]methanol containing sodium [D3]-
methoxide at room temperature, 11 was converted into exo-
bicyclo[1.1.0]but-2-ylcarbinyl [D3]methyl ether (16) in good
yield (Scheme 6).
following characteristic coupling constants: J1,2 ) J2,3 ) 3.3
Hz for endo-9 and 0.9 Hz for exo-9; JC2,H2 ) 150.7 Hz for
endo-9 and about 170 Hz for exo-9.
Conversion of the Alcohols 9 into the Sulfonates 10 and
11. With regard to the preparation of sulfonates from the
alcohols 9, we anticipated the same difficulties as experienced
previously in the case of related compounds.2,3 Thus, a mixture
of endo-9 and triethylamine in CDCl3 was treated with meth-
anesulfonyl chloride at -36 °C. The desired mesylate 10 could
be observed by NMR spectroscopy of the reaction mixture
without workup. However, the persistence of 10 was found to
be limited. On standing at room temperature for a day, 40% of
the amount of 10 in the NMR sample converted into cyclopent-
3-en-1-yl mesylate (12) and a second 4-substituted cyclopentene
in the ratio of 2:1. Whereas 12 was the product of the internal
return of the ion pair from 13 and the mesylate ion formed in
the heterolytic dissociation of 10 (Scheme 5), the second product
resulted from the collapse of 13 with another nucleophile of
the sample, for example, triethylamine, chloride, or unintention-
ally present water. The consumption of 10 was complete after
9 days at room temperature.
An analogous experiment to prepare the mesylate of exo-9
failed, most probably because of the acidification of the mixture
brought about by the formation of triethylammonium chloride,
which is supposed to catalyze irreversible reactions of the
bicyclo[1.1.0]butane system of exo-9 and possibly its mesylate
as well. We then turned to a method that avoids acids entirely
and, thus, deprotonated exo-9 with sodium hydride prior to the
addition of tosyl chloride to the reaction mixture. Indeed, the
desired tosylate 11 was produced and obtained as a 4:1 mixture
with 1-butyl tosylate after the workup. At variance with the
behavior of 10, an NMR sample of 11 remained unchanged at
room temperature. The identity of the sulfonates 10 and 11 was
clearly established by their NMR spectra. Whereas the multi-
plicities of the signals were very similar to those of the alcohols
9, several chemical shifts showed the typical changes expected
for the conversion of an alcohol into its sulfonate.3,4
Attempts to attain a hydrolysis product of 11 were to no avail.
Whereas the consumption of 11 in aqueous acetone of various
concentrations in the presence of sodium bicarbonate or
triethylamine seemed to proceed surprisingly slowly, a product
could not be identified. The monitoring by NMR spectroscopy
of the fate of 11 in 90% aqueous [D6]acetone containing
triethylamine then revealed the formation of (exo-bicyclo[1.1.0]-
but-2-ylcarbinyl)triethylammonium tosylate (17; Scheme 6). At
room temperature, 11 was consumed to the extent of 71% within
118 days, with the yield of 17 amounting to 41%. The identity
of 17 was determined on the basis of the similarity of its NMR
spectroscopic data with those of 11 and, in particular, by the
highly characteristic splitting of certain signals as a result of
the scalar coupling of the respective nuclei with the 14N nucleus.9
Being present as an impurity, 1-butyl tosylate underwent
analogous reactions as 11, that is, it was transformed to 1-butyl
[D3]methyl ether and 1-butyltriethylammonium tosylate, re-
spectively. Throughout these reactions, the ratio of the concen-
trations of 11 and 1-butyl tosylate remained constant. In the
case of the methanolysis, a variation of the reaction rate was
qualitatively observed when different concentrations of the
substrates were utilized. These findings are in line with SN2
processes for solvolyses of 11 and at variance with the
intermediacy of the exo-bicyclo[1.1.0]but-2-ylcarbinyl cation
(exo-18; Scheme 6).
In summary, the products 12, 14, and 15 on one hand and 16
and 17 on the other are in harmony with the results of the
solvolyses of endo,endo- and exo,exo-4 (Scheme 2) and
emphasize the discrepancy between them and the formation of
the common product 3 on the solvolysis of endo- and exo-2
(Scheme 1).
Kinetic Studies. Rate constants were obtained for the
solvolyses in methanol (Table 1) and acetone/water (Table 2).
A scaled-down procedure for the in situ preparation of mesylates
was tested for the known solvolyses of 1-adamantyl mesylate
in methanol (monitored conductometrically).10 The method is
2. Solvolysis of endo-Bicyclo[1.1.0]but-2-ylcarbinyl Mes-
ylate (10) and exo-Bicyclo[1.1.0]but-2-ylcarbinyl Tosylate
(11). Product Studies. As the thermal rearrangement of 10 to
cyclopent-3-en-1-yl mesylate (12) already indicated, the het-
erolytic dissociation of 10 gives rise to the cation 13 (Scheme
5). This was confirmed by the dissolution of 10 in 75% acetone/
water as well as in [D4]methanol containing sodium [D3]-
methoxide at room temperature. In both cases, 12 emerged as
one of the products, while the second was cyclopent-3-en-1-ol
(14) and cyclopent-3-en-1-yl [D3]methyl ether (15), respectively.
(9) Berger, S.; Braun, S.; Kalinowski, H.-O. NMR-Spektroskopie von
Nichtmetallen. 15N NMR-Spektroskopie, Band 2; Georg Thieme Verlag:
Stuttgart, 1992.
1020 J. Org. Chem., Vol. 71, No. 3, 2006

Unsubstituted Bicyclo[1.1.0]butyl-2-carbinyl Cations
TABLE 1. Rate Constants (k) for Solvolyses of Sulfonates in
Methanol Containing Triethylamine
CHART 1
substrate
T (°C)
NEt3 (M)
k (s-1
)
1-AdOMsa,b
25.0
25.0
40.0
40.0
40.0
40.0
40.0
40.0
25.0
c
c
c
(2.71 ( 0.02) × 10-4
(1.42 ( 0.04) × 10-4
(8.65 ( 0.17) × 10-4
(8.72 ( 0.09) × 10-4
(2.95 ( 0.2) × 10-6
(1.11 ( 0.08) × 10-6
(3.32 ( 0.13) × 10-6
(1.08 ( 0.02) × 10-6
≈6.0 × 10-7
endo-OMs (10)
endo-OMs (10)d,e
endo-OMs (10)
exo-OTs (11)f
1-BuOTsg
0.01
0.01
0.01
0.003
0.003
exo-OTs (11)g
1-BuOTsg
exo-OTs (11)h,i
supporting an SN2 mechanism (see also the above discussion
of Scheme 6). Although low concentrations of the nucleophile
triethylamine were present in the solutions used for the kinetic
studies, very similar rate constants were obtained for metha-
nolyses in the presence 0.003 and 0.01 M triethylamine (Table
1), so the data should refer to the pseudo-first-order rate
constants for methanolysis. However, the rate constants for the
corresponding SN1 solvolyses via the cation exo-18 would be
lower.
a Determined conductometrically in duplicate; errors shown are average
deviations. b Reference 10: k ) (2.81 ( 0.02) × 10-4
.
c The solution
contained a small excess of triethylamine, carried through from the in situ
preparation of the mesylate. d Determined conductometrically in triplicate.
e Activation parameters: ∆Hq ) 21.8 kcal mol-1; ∆Sq ) -3.2 cal mol-1
K-1 f The rate constant is an average of one conductometric and one kinetic
.
run monitored by HPLC. g The rate constant was obtained from one HPLC
kinetic run for a mixture of 1-butyl tosylate and 11. h Estimated from data
at 40 °C, assuming ∆Sq ) -20 cal mol-1 K-1 (for an SN2 alcoholysis of
a benzenesulfonate, see ref 12). i Two reactions monitored by 1H NMR
(see experimental) at 22 °C in the presence of about 0.3 M methoxide gave
higher rate constants (approximate half-lives were 9-24 h, corresponding
to k ≈ 10-5 s-1 ).
Similar considerations apply to the experiments with acetone/
water, with added triethylamine. As conditions for preparative
studies (Scheme 6) involve about 100-fold higher concentrations
of tosylate 11 and triethylamine than conditions for kinetic
studies (Table 2), the formation of the salt 17 is less likely in
the kinetic studies. The kinetic monitoring by HPLC with UV
detection reveals only the formation of p-toluenesulfonic acid
as product, but the effect of changes in the nucleophile
concentrations can be illustrated using published data for methyl
tosylate; in methanol at 25 °C, second-order rate constants for
the reaction with triethylamine are over 20 000-fold greater than
calculated values for methanolysis,14 but 0.003 M triethyl-
amine is 10 000-fold more diluted than the methanol solvent.
As competing SN2 reactions should show a rate/product cor-
relation, methanolysis and aminolysis are competitive reactions
in very dilute amine solutions, even for the less-hindered methyl
tosylate. Consequently, methanolysis of 11 should proceed under
kinetic conditions in addition to aminolysis. Another indication
of the consistency of the data is the similarity of the rate
constants for both methanolysis and hydrolysis (in 60% acetone/
water) of 11, comparing again with an SN2 mechanism for
methyl tosylate, the rate constants of which are 1.06 × 10-5
and 1.38 × 10-5 s-1 in methanol and 60% acetone/water,
respectively, at 50 °C.15
In contrast, the rates of solvolyses of the endo-mesylate 10
increase 70-fold from methanol to 40% acetone/water, showing
a significant dependence on the solvent ionizing power but less
than that for the solvolyses of cyclopropylcarbinyl substrates.16
Rates for 10 are similar to those of the solvolyses of the
corresponding monocyclic compound, that is, cyclopropylcarbi-
nyl mesylate (19, Chart 1). In 60% acetone/water at 25 °C, the
rate constant for 19 is 5.25 × 10-3 s-1,3 that is three times as
large as that of 10 (Table 2). A cationic (SN1) mechanism in
which rearrangement accompanies ionization accounts for the
kinetic data and also for the products (Scheme 5). As 11 and
10 react by different mechanisms, the exo/endo rate ratio is
solvent dependent and varies from 1:300 in methanol at 40 °C
(Table 1) to 1:3600 in 60% acetone/water at 25 °C (Table 2).
TABLE 2. Rate Constants (k) for Solvolyses of Sulfonates in
Acetone/Water (% v/v) Containing Triethylamine at 25.0 °C
substrate
% v/v
NEt3 (M)
k (s-1
)
endo-OMs (10)a
endo-OMs (10)a,c
exo-OTs (11)d
1-BuOTsd
60
40
60
60
b
b
(1.75 ( 0.04) × 10-3
(1.00 ( 0.03) × 10-2
(4.82 ( 0.12) × 10-7
(1.62 ( 0.06) × 10-7
0.003
0.003
a Refer to Table 1, footnote a. b Refer to Table 1, footnote c. c An
acetonitrile solution of mesylate was injected. d Refer to Table 1, footnote
g.
suitable for the solvolyses of mesylates, which are not as reactive
as those we studied earlier.2,11 The reactions of the exo-tosylate
11 in the presence of 1-butyl tosylate were also monitored by
HPLC; as separate peaks were then obtained for the two
tosylates, both rate constants were obtained (Tables 1 and 2).
The exo-tosylate 11 reacted three times as fast as 1-butyl
tosylate (three measurements, Tables 1 and 2). Normally,
â-branching reduces the rates of SN2 reactions, for example, in
ethanol at 50 °C, iso-butyl benzenesulfonate reacts 14-fold
slower than 1-propylbenzenesulfonate.12 These results provide
some evidence of enhanced rates for the solvolyses of 11.
Relative rates of solvolyses of geometrically constrained cy-
clopropylcarbinyl systems depend strongly on the angle of
rotation of the cyclopropane group against the planar cationic
subunit.13 Rates may be strongly enhanced, but in a constrained
perpendicular conformation, rates may even be reduced by a
factor of about 200.13 Although rotation in 11 is not prevented,
the preferred direction of attack by a nucleophile may lead to
a transition state of nonoptimal conformation having a partial
positive charge on the carbinyl carbon atom.
Reactions of 11 in methanol were faster in the presence of
higher concentrations of methoxide base (Table 1, footnote i),
(10) Bentley, T. W.; Carter, G. E. J. Org. Chem. 1983, 48, 579-584.
(11) Bentley, T. W.; Kirmse, W.; Llewellyn, G.; So¨llenbo¨hmer, F. J.
Org. Chem. 1990, 55, 1536-1540.
(12) Laughton, P. M.; Robertson, R. E. Can. J. Chem. 1955, 33, 1207-
1215.
(13) Rhodes, Y. E.; DiFate, V. G. J. Am. Chem. Soc. 1972, 94, 7582-
7583.
(14) Pearson, R. G.; Songstad, J. J. Org. Chem. 1967, 32, 2899-2900.
(15) Schadt, F. L.; Bentley, T. W.; Schleyer, P. v. R. J. Am. Chem. Soc.
1976, 98, 7667-7674.
(16) Kevill, D. N.; Abduljaber, M. H. J. Org. Chem. 2000, 65, 2548-
2554.
J. Org. Chem, Vol. 71, No. 3, 2006 1021

Bentley et al.
FIGURE 1. Structure and relative free energies computed for the C5H7 cations 13, 18, 25, and 26. The label TS characterizes a species as a
transition state.
For the reasons discussed above, these ratios overestimate the
SN1 reactivity of 11. Consequently, the 11/10 (exo/endo) rate
ratio for SN1 reactions must be even less than the above
estimates, whereas the exo/endo rate ratio is 2 for the solvolyses
of 2.1 The reasons for this major difference are explored
theoretically in the next section.
on one side and C1 and C3 on the other, having 80 and 70% p
character, respectively.19 Thus, the corresponding electrons are
less suitable for participation than the respective ones of 23,
causing the slow solvolysis of the exo-tosylate 11. The same
effect should be operative in the case of the endo-mesylate 10,
but there the endo-bicyclo[1.1.0]but-2-ylcarbinyl cation is
bypassed during solvolysis as a result of the concomitant
Wagner-Meerwein rearrangement [see endo-18(TS) in Figure
1], which brings about a significant release of strain energy and,
therefore, almost compensates the rate-retarding effect of the
participating bonds in comparison with 19. Such compensation
is not possible for the exo-tosylate 11 because of the inability
of the cation exo-18 to undergo a conventional Wagner-
Meerwein rearrangement (see next section).
The introduction of a second OMs or OTs group facilitated
our earlier work by deactivating the substrates.2,4 The magnitude
of the effect in a relatively flexible substrate can be calculated
from the solvolysis rates of the endo-mesylate 10 and endo,endo-
4. In acetone/water at 25 °C, the ratio amounts to 21 and 31
(60%, 1.75 × 10-3 s-1 versus 8.2 × 10-5; 40%, 1.00 × 10-2
versus 3.24 × 10-4; data from Table 2 and ref 4) and is, thus,
20- to 30-fold less than we expected previously.4
3. Calculations of the Stability and the Rearrangements
of the Bicyclo[1.1.0]butyl-2-carbinyl Cations. Computational
Details. All structures were optimized by means of analytical
gradients in combination with the B3LYP functional20 in the
TZVP (triple zeta valence quality with polarization functions)21
basis set, which for carbon represents an (11s6p1d) atomic
orbital (AO) basis in a [5s3p1d] contraction and for hydrogen
a (10s1p) AO basis in a [4s1p] contraction. Solvent effects were
estimated with the aid of COSMO (conductor-like screening
model)22 with a dielectric constant of ꢀ ) 30 to simulate the
acetone solvent. Energy minima and transition states were
The difference between the rate of 10 and the rate of 11 as
well as the fact that both solvolyse more slowly than 19 calls
for an explanation. A clue is provided by the reactivity of the
bicyclo[1.1.0]but-1-ylcarbinyl esters 21 and 22 (Chart 1). The
tosylate 21 could not be obtained as a result of its rapid
rearrangement to 3-methylenecyclobutyl tosylate,17 but the
intermediacy of the bicyclo[1.1.0]but-1-ylcarbinyl cation (24,
Chart 1) is highly likely because 24 was observed directly in a
non-nucleophilic medium.18 Estimated roughly, the rate of
solvolysis of the p-nitrobenzoate 22 is 1000 times greater than
that of cyclopropylcarbinyl p-nitrobenzoate (20).17 This differ-
ence is caused by the availability for participation of the
electrons of the C-C bonds adjacent to the cationic center. The
more p character these bonds have, the more stabilized is the
cation and, hence, the higher the rate of solvolysis.
Whereas the p character of these bonds amounts to 80% in
the cyclopropylcarbinyl cation (23), it is much larger for the
bond between the bridgehead carbon atoms of 24, which is
considered to be formed from electrons in the almost pure p
orbitals.19 A consequence of this high p character of the central
bicyclo[1.1.0]butane bond is that the lateral bonds are formed
from electrons in orbitals with less p character than those of
the C-C bonds of cyclopropane. Being the participating ones
in the bicyclo[1.1.0]but-2-ylcarbinyl cations (18, Scheme 6 and
Figure 1), such lateral bonds emerge from the orbitals of C2
(17) Wiberg, K. B.; Lampman, G. M.; Cuila, R. P.; Connor, D. S.;
Schertler, P.; Lavanish, J. Tetrahedron 1965, 21, 2749-2769.
(18) Wiberg, K. B.; McMurdie, N. J. Org. Chem. 1993, 58, 5603-5604.
(19) Wiberg, K. B. In The Chemistry of the Cyclopropyl Group;
Rappoport, Z., Ed.; Wiley: Chichester, England, 1987; Part 1, pp 1-26.
(20) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(21) Scha¨fer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100,
5829-5835.
1022 J. Org. Chem., Vol. 71, No. 3, 2006

Unsubstituted Bicyclo[1.1.0]butyl-2-carbinyl Cations
TABLE 3. Gas-Phase Energies (∆E, E ) 1), Gas-Phase Free
Energies (∆G, E ) 1), and Energies in Solution (∆E, E ) 30) of
Various C5H7 Cations, Relative to the Corresponding Values of 13a
predicted a folded structure as the ground state and a planar
five-membered ring (classical cyclopent-3-en-1-yl cation) as the
transition state for the interconversion of the degenerate envelope
forms. The latter was calculated to be more stable by 4.5 kcal
mol-1 than the planar isomer after the correction for the zero-
point energy.29
The finding that endo-18(TS) and endo-18′(TS) reorganize
to furnish 13 without an activation barrier corroborates our
previous assumption of the Wagner-Meerwein rearrangement
as proceeding simultaneously with the dissociation during most
solvolyses of corresponding bicyclo[1.1.0]but-2-ylcarbinyl sub-
strates. Thus, derivatives of endo-18(TS), which might be
thought to be generated from endo,endo-4, tricyclo[3.1.0.02,6]-
hex-3-yl sulfonates, and tricyclo[4.1.0.02,7]hept-3-yl esters,
cannot be considered as intermediates.2-4 Only in the case of
tricyclo[4.1.0.02,7]hept-4-en-3-yl p-nitrobenzoate, which was the
sole substrate to furnish nonrearranged solvolysis products to
some extent, the intermediacy of the tricyclo[4.1.0.02,7]hept-4-
en-3-yl cation, which is an allyl cation bridged by a bicyclo-
[1.1.0]butane system, seems possible at least.3
∆E
∆G298
(B3LYP/TZVP;
ꢀ ) 1)
∆E
(B3LYP/TZVP;
ꢀ ) 1)
(B3LYP/TZVP;
ꢀ ) 30)
cation
endo-18(TS)
endo-18′(TS)
exo-18
exo-18(TS)
25(TS)
40.7
65.3
40.6
59.5
51.8
4.1
42.7
66.2
41.1
57.0
49.0
4.6
41.2
64.7
41.3
26
3.6
a See Figure 1; in kcal mol-1
.
checked by frequency calculations. All computations were
performed with the Turbomole program package,23 which also
allowed the determination of thermodynamic corrections by the
standard implementation.
endo- [endo-18(TS)] and exo-Bicyclo[1.1.0]but-2-ylcarbinyl
Cations (exo-18). The gas-phase energies of the C5H7 cations
18, 25, and 26, relative to the value of 13, are collected in Table
3. In addition, the corresponding ∆G values at 298 K are
contained as well as the energies in solution (ꢀ ) 30) of species
18 and 26, again relative to the corresponding data of 13. Solvent
effects are not discussed specifically because they do not seem
to be of particular importance with regard to the energy
differences, as calculations have shown for the cyclopropyl-
carbinyl cation and its isomers.24 Also, thermodynamic correc-
tions do not cause significant changes in this respect. Figure 1
displays the structures of 13, 18, 25, and 26, which are grouped
according to their relative free energies and their fates on
rearrangement.
Two conformations of the endo-bicyclo[1.1.0]but-2-yl cation
have been calculated. As expected, in analogy to the most stable
form of the cyclopropylcarbinyl cation,24-26 the bisected
structure [endo-18(TS)] is much lower in energy than the
perpendicular one ([endo-18′(TS)]), with the free energy dif-
ference being 23.5 kcal mol-1. Neither of these two cations is
an energy minimum, as both undergo a [1,2]-C migration
without an activation barrier to give the nonclassical cyclopent-
3-en-1-yl cation (13). In fact, endo-18(TS) and endo-18′(TS)
are transition states for automerization pathways of 13, which
have no significance, however, because the isomerization of 13
leading to the cyclopent-2-en-1-yl cation, brought about by a
[1,2]-H shift, has a much lower activation barrier.27 The
nonclassical nature of 13, being a bishomocyclopropenyl cation,
has previously been demonstrated by calculations28,29 that, hence,
In contrast to endo-18(TS), the exo-bicyclo[1.1.0]but-2-
ylcarbinyl cation in the bisected conformation (exo-18) indeed
represents an energy-minimum structure, as expected for a
cyclopropylcarbinyl cation;25 its free energy (41.1 kcal mol-1
,
Table 3, Figure 1) is very similar to that of endo-18(TS). The
calculations show that the most favorable isomerization of exo-
18 leads also to 13. However, the mechanism of the process is
not a simple Wagner-Meerwein rearrangement, but one that
occurs in two stages. Its transition state has a structure
resembling that of the 2-vinylcycloprop-1-yl cation [25(TS)],
whose free energy is 7.9 kcal mol-1 greater than that of exo-
18. In line with this result, we have pointed out previously that
the inability of the cation conceivably generated from exo,exo-4
(Scheme 2) to undergo a [1,2]-C shift might be the reason for
the exclusive formation of unrearranged products such as 6, even
if SN2 reactions of exo,exo-4 could not be rigorously ruled out.4
The substitution products, 16 and 17, from the solvolyses of
11 support this interpretation.
There is a second, less favorable, pathway for exo-18 to
release most of its considerable strain energy, namely, a [1,2]-
H migration to the cationic center giving rise to 26, which also
is a homoaromatic species, that is, the nonclassical 2-methyl-
cyclobutenyl cation. The transition state [exo-18(TS)] is the
slightly twisted perpendicular conformation of the exo-bicyclo-
[1.1.0]but-2-ylcarbinyl cation. By an extensive study using
theoretical methods, the homoaromatic nature of the unsubsti-
tuted cyclobutenyl cation, the parent of 26, has been demon-
strated.30
Introduction of Methyl Groups into the Bridgehead
Positions of the Bicyclo[1.1.0]but-2-ylcarbinyl Cations. The
replacement of the hydrogen atoms of the positions 1 and 3 of
all species 18 by methyl groups has a dramatic effect on the
shape of the potential energy surface. While exo-18 represents
a local minimum and has to surmount a barrier of 7.9 kcal mol-1
to rearrange to 13, its methylated counterpart 27(TS) converts
into 28 (Scheme 7), which is the 1,3-dimethyl derivative of 13,
without any barrier. Because the positive charge of 25(TS)
largely resides at a carbon atom whose hydrogen atom is
(22) Klamt, A.; Schu¨u¨rmann, G. J. Chem. Soc., Perkin Trans. 2 1993,
799-805.
(23) Ahlrichs, R.; Ba¨r, M.; Baron, H.-P.; Bauernschmitt, R.; Bo¨cker, S.;
Deglmann, P.; Ehrig, M.; Eichkorn, K.; Elliott, S.; Furche, F.; Haase, F.;
Ha¨ser, M.; Horn, H.; Ha¨ttig, C.; Huber, C.; Huniar, U.; Kattannek, M.;
Ko¨hn, A.; Ko¨lmel, C.; Kollwitz, M.; May, K.; Ochsenfeld, C.; O¨ hm, H.;
Scha¨fer, A.; Schneider, U.; Sierka, M.; Treutler, O.; Unterreiner, B.; von
Arnim, M.; Weigend, F.; Weis, P.; Weiss, H. TURBOMOLE, version 5.6;
Universita¨t Karlsruhe: Germany, since 1988.
(24) Casanova, J.; Kent, D. R., IV; Goddard, W. A., III; Roberts, J. D.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 15-19.
(25) Saunders, M.; Laidig, K. E.; Wiberg, K. B.; Schleyer, P. v. R. J.
Am. Chem. Soc. 1988, 110, 7652-7659.
(26) Wiberg, K. B.; Shobe, D.; Nelson, G. L. J. Am. Chem. Soc. 1993,
115, 10645-10652.
(27) Olah, G. A.; Prakash, G. K. S.; Rawdah, T. N.; Whittaker, D.; Rees,
J. C. J. Am. Chem. Soc. 1979, 101, 3935-3939.
(29) Szabo, K. J.; Kraka, E.; Cremer, D. J. Org. Chem. 1996, 61, 2783-
2800.
(30) Sieber, S.; Schleyer, P. v. R.; Otto, A. H.; Gauss, J.; Reichel, F.;
Cremer, D. J. Phys. Org. Chem. 1993, 6, 445-464.
(28) Schleyer, P. v. R.; Bentley, T. W.; Koch, W.; Kos, A. J.; Schwarz,
H. J. Am. Chem. Soc. 1987, 109, 6953-6957.
J. Org. Chem, Vol. 71, No. 3, 2006 1023

Bentley et al.
SCHEME 7. Schematic Representation of the Relationship
between the C5H5(CH3)2 Cations 27(TS) and 28
more stable than the exo-1,3-dimethylbicyclo[1.1.0]but-2-yl-
carbinyl cation, whereby this cation adopts the character of a
transition state [27(TS)]. This finding explains the products and
the rates of the solvolysis of the diastereomeric 1,3-dimethylbi-
cyclo[1.1.0]but-2-ylcarbinyl tosylates (2)1 in comparison with
the results obtained from the diastereomeric pair endo,endo-
(endo,endo-4) and exo,exo-bicyclo[1.1.0]but-2,4-dimethanol dimes-
ylate (exo,exo-4)4 as well as 10 and 11.
replaced by a methyl group, a substantial stabilization occurs.
Such an effect is operative neither on going from exo-18 to
27(TS) nor on going from 13 to 28, but both of these transitions
are influenced in very much the same way, as is indicated by
the energy differences between exo-18 and 13 on one hand and
between 27(TS) and 28 on the other, which were calculated to
be 41.1 (Figure 1, Table 3) and 42.5 kcal mol-1 (Scheme 7),
respectively. Additionally, it can be assumed that the 1,3-
dimethyl derivative of endo-18(TS) is also a transition state and
retains an energy relative to that of 28 that is similar to the
energy difference between endo-18(TS) and 13.
Clearly, this result explains that the tosylates endo- and exo-2
can furnish the common product 3,1 which emerges from the
trapping of the cation 28 by water, and that both substrates may
react at about the same rate. However, the 1,3-dimethylbicyclo-
[1.1.0]but-2-ylcarbinyl cations [see 27(TS)] cannot be consid-
ered as intermediates because the rearrangements should occur
concomitant with the heterolytic dissociation.
Experimental Section
General Methods. See ref 4.
erythro- (erythro-8) and threo-2′,2′-Dibromocycloprop-1′-
yloxirane (threo-8). A solution of dimethyldioxirane8 (2.50 mmol,
31.25 mL of 0.08 M in acetone) was added with stirring to 1,1-
dibromo-2-vinylcyclopropane7 (500 mg, 2.21 mmol), which was
kept at 0 °C under nitrogen, within 50 min. Stirring was continued
for 2 h at 0 °C and then overnight at room temperature. The mixture
was concentrated in vacuo to give a light yellow oil (400 mg, 75%),
which was shown by NMR spectroscopy to consist of a virtually
pure 1:1 mixture of erythro- and threo-8. Flash chromatography
[SiO2; light petroleum ether (bp 30-50 °C)/tert-butyl methyl ether,
20:1] afforded pure threo-8 (150 mg, 28%) and pure erythro-8 (170
mg, 32%). threo-8: Rf ) 0.47, colorless crystals, mp 41-43 °C.
1H NMR (400 MHz, CDCl3): δ 1.63 (t, J ) 7.3 Hz, 1H), 1.76
(dd, J ) 10.3, 7.2 Hz, 1H), 1.85 (ddd, J ) 10.3, 7.4, 3.8 Hz, 1H),
2.64 (dd, J ) 5.1, 2.6 Hz, 1H), 2.89 (dd, J ) 5.1, 3.9 Hz, 1H),
3.09 (td, J ) 3.8, 2.6 Hz, 1H). 13C NMR (101 MHz, CDCl3): δ
23.4 (s), 24.8 (t), 31.8 (d), 45.9 (t), 50.7 (d). Anal. Calcd for C5H6-
Br2O: C, 24.83; H, 2.50. Found: C, 25.06; H, 2.44. erythro-8: Rf
) 0.34, colorless oil. 1H NMR (400 MHz, CDCl3): δ 1.54 (t, J )
7.4 Hz, 1H), 1.61 (br dt, J ) 10.1, 7.2 Hz, 1H), 1.84 (dd, J ) 10.1,
7.2 Hz, 1H), 2.70 (dd, J ) 4.9, 2.6 Hz, 1H), 2.87 (br t, J ) 4.4 Hz,
1H), 2.95 (br ddd, J ) 7.0, 3.7, 2.6 Hz, 1H). 13C NMR (101 MHz,
CDCl3): δ 23.8 (s), 26.0 (t), 32.0 (d), 46.1 (t), 53.2 (d). Anal. Calcd
for C5H6Br2O: C, 24.83; H, 2.50. Found: C, 24.97; H, 2.53.
endo-Bicyclo[1.1.0]butane-2-methanol (endo-9). n-Butyllithium
(5.00 mmol, 3.31 mL of 1.51 M in hexane) was added dropwise
within 35 min to a stirred solution of threo-8 (550 mg, 2.27 mmol)
in anhydrous diethyl ether (40 mL), which was kept at -75 °C
under nitrogen. Stirring was continued for 1.5 h at -75 to -70 °C
and then for 1 h at 20 °C. The mixture was treated with water (5
mL), the resulting layers were separated, the aqueous phase was
extracted with diethyl ether (3 × 10 mL), and the combined organic
phases were washed with brine (2 × 25 mL). After drying with
Na2SO4, the solution was concentrated in vacuo (500-250 mbar)
at 20 °C. A 1H NMR spectrum of the residue indicated the presence
of endo-9, 1-bromobutane, and 1-butanol. The desired product,
endo-9, was enriched by a slow evaporation of the volatile
components of the mixture in vacuo (250-0.07 mbar) at 20 °C
and their condensation in cooled receivers. The first fractions
consisted mainly of 1-bromobutane and 1-butanol, whereas the last
fractions contained endo-9 as the major component. Out of a number
of experiments, the best fraction obtained was a colorless liquid
(66 mg) composed of endo-9 (32% yield) and 1-butanol in the ratio
Conclusion
Motivated by previous results that apparently contradicted
each other,1,4 we undertook to approach the problem by
experimental and theoretical studies of the unsubstituted dia-
stereomeric bicyclo[1.1.0]but-2-ylcarbinyl cations. The experi-
ments included the development of a synthetic route to the two
bicyclo[1.1.0]butane-2-methanols (9), their conversion into the
mesylate 10 and the tosylate 11 in the case of the endo- and
exo-isomer, respectively, and the determination of the products
and the rates of the solvolysis reactions of 10 and 11. Both
mesylate 10 and tosylate 11 underwent solvolysis at smaller
rates than cyclopropylcarbinyl mesylate (19), whereas bicyclo-
[1.1.0]but-1-ylcarbinyl esters had previously been shown to react
faster than the corresponding cyclopropylcarbinyl esters.17 These
phenomena are readily explained on the basis of the p character
of the C-C bonds that can exert neighboring group participation
on heterolytic dissociation.
The quantum chemical calculations revealed that the endo-
bicyclo[1.1.0]but-2-ylcarbinyl cation is only a transition state
[endo-18(TS)] for an automerization of the nonclassical cyclo-
pent-3-en-1-yl cation (13), whereas the exo isomer corresponds
to a local energy minimum (exo-18). In line with these
predictions, the solvolyses of the endo-mesylate 10 gave rise
to 4-substituted cyclopentenes exclusively, thus demonstrating
that the Wagner-Meerwein rearrangement occurred concomi-
tant with the heterolytic dissociation. The outcome of the
solvolyses of the exo-tosylate 11 is formally also in agreement
with the computational result, as only products with complete
retention of the carbon atom skeleton were observed. However,
the rate studies do not support the intermediacy of the cation
exo-18 but favor an SN2 process.
1
of about 15:1 and only traces of other impurities. H NMR (400
MHz; C6D6; the chemical shifts in CDCl3 are given in brackets):
δ 0.89 [1.15] (t, J ) 5.9 Hz, 1H), 1.14 [1.52] (td, J ) 3.3, 1.7 Hz,
2H), 1.25 [1.44] (qq, J ) 1.7, 0.5 Hz, 1H), 1.48 [1.84] (qd, J )
3.3, 1.7 Hz, 1H), 2.38 [2.68] (tqd, J ) 7.2, 3.3, 0.6 Hz, 1H), 3.21
[3.36] (br dd, J ) 7.2, 5.9 Hz, 2H). 13C NMR (101 MHz; C6D6;
the chemical shifts in CDCl3 are given in brackets): δ 1.7 [1.4]
(dsext, JC,H ) 201.3, 3.5 Hz, 2C), 30.3 [30.2] (dddt, JC,H ) 170.6,
151.4, 12.7, 4.2 Hz, 1C), 49.7 [49.4] (ddquint, JC,H ) 150.7, 14.2,
3.4 Hz, 1C), 56.0 [56.3] (br t, JC,H ) 141.5 Hz, 1C). MS (EI, 70
eV, %) m/z: M+ 84 (11), 83 (35), 69 (14), 67 (10), 56 (33), 55
(100), 54 (11), 53 (24), 43 (15), 41 (44), 39 (35); the intensity of
Most interestingly, the introduction of methyl groups in
positions 1 and 3 of exo-18 is predicted to have a dramatic effect
on the transition state that separates exo-18 from 13. It becomes
1024 J. Org. Chem., Vol. 71, No. 3, 2006

Unsubstituted Bicyclo[1.1.0]butyl-2-carbinyl Cations
several signals may have a contribution from 1-butanol. HRMS
(EI, 70 eV) m/z: [M+ - H] calcd for C5H7O, 83.0497; found,
83.0495.
finally with water (2 × 5 mL). Drying of the organic layer with
K2CO3/Na2SO4 and its concentration in vacuo furnished a colorless
oil (115 mg), shown by NMR spectroscopy to be a 4:1 mixture of
1
11 (57%) and 1-butyl tosylate. H NMR (400 MHz, CDCl3): δ
exo-Bicyclo[1.1.0]butane-2-methanol (exo-9). According to the
procedure for the preparation of endo-9, exo-9 (28%) was obtained
from erythro-8 (1.25 g, 5.17 mmol) in the best case as a 4:1 mixture
(150 mg, colorless liquid) with 1-butanol, containing only traces
of other impurities. 1H NMR (400 MHz; CDCl3; the chemical shifts
in C6D6 are given in brackets): δ 0.60 [0.53] (qd, J ) 1.2, 0.9 Hz,
1H), 1.16 [1.05] (tdtd, J ) 5.8, 1.2, 0.9, 0.4 Hz, 1H), 1.28 [0.95]
(br, 1H), 1.45 [1.18] (ddd, J ) 2.9, 1.2, 0.9 Hz, 2H), 1.50 [1.27]
0.56 (br quint, J ) 1.0 Hz, 1H), 1.09 (tm, J ) 6.3 Hz, 1H), 1.44
(td, J ) 2.9, 0.8 Hz, 1H), 1.54 (ddd, J ) 2.9, 0.9, 0.5 Hz, 2H),
2.45 (s, 3H), 3.93 (d, J ) 6.3 Hz, 2H), 7.34 (m, 2H), 7.79 (m, 2H).
13C NMR (101 MHz, CDCl3): δ 0.5 (dquint, JC,H ) 207.4, 3.9
Hz, 2C), 21.6 (qt, JC,H ) 127.3, 4.3 Hz, 1C), 27.7 (ddm, JC,H
170, 154 Hz, 1C), 40.0 (dm, JC,H ≈ 170 Hz, 1C), 69.7 (tm, JC,H
150 Hz, 1C), 127.9 (dm, JC,H ≈ 165 Hz, 2C), 129.8 (dm, JC,H
162 Hz, 2C), 133.2 (m, 1C), 144.7 (m, 1C).
)
≈
≈
(tdd, J ) 2.9, 0.9, 0.4 Hz, 1H), 3.55 [3.25] (t, J ) 5.8 Hz, 2H). 13
C
NMR (101 MHz; C6D6; the chemical shifts in CDCl3 are given in
brackets): δ -0.3 [-0.7] (dquint, JC,H ) 203.7, 3.8 Hz, 2C), 28.2
[28.1] (ddq, JC,H ) 169.1, 153.0, 4.2 Hz, 1C), 46.0 [45.5] (ddm,
JC,H ≈ 170, 13 Hz, 1C), 62.1 [62.5] (tm, JC,H ≈ 141 Hz, 1C). MS
(EI, 70 eV, %) m/z: M+ 84 (3), 83 (9), 56 (100), 55 (39), 44 (18),
43 (48), 42 (27), 41 (66), 40 (57), 39 (17); the intensity of several
signals may have a contribution from 1-butanol. HRMS (EI, 70
eV) m/z: [M+ - H] calcd for C5H7O, 83.0497; found, 83.0496.
endo-Bicyclo[1.1.0]but-2-ylcarbinyl Mesylate (10). A stirred
solution of endo-9 (14 mg, 0.17 mmol), containing some 1-butanol,
and dry triethylamine (25 mg, 0.25 mmol) in anhydrous dichlo-
romethane (0.5 mL) was kept at -36 °C under nitrogen and treated
dropwise with methanesulfonyl chloride (19 mg, 0.17 mmol) in
anhydrous dichloromethane (0.2 mL) within 8 min. Stirring was
continued for 90 min at -35 to -30 °C, while the progress of the
reaction was monitored by TLC on basic Al2O3 (activity I) with
petroleum ether (bp 30-50 °C)/tert-butyl methyl ether, 5:4. The
mixture was then allowed to warm to -5 °C and, after the addition
of 2 mL of cold dichloromethane, was washed quickly with ice-
cold water (2 × 3 mL). The organic phase was dried with K2CO3/
Na2SO4 and then concentrated in vacuo at 0-5 °C. The NMR
spectra of the residue (20 mg of a light yellow oil) were taken at
27 °C and showed the presence of the mesylate 10, 1-butyl mesylate,
dichloromethane, and water. The sample was kept at -30 °C until
Solvolysis of 10 in 75% Aqueous Acetone. Triethylamine (50
mg, 0.49 mmol) and the mesylate 10 (20 mg, 0.12 mmol),
containing some 1-butyl mesylate, were dissolved in a minimum
quantity of acetone (1.5 mL). After the addition of water (0.5 mL),
the mixture was stirred at 22 °C for 18 h. Then most of the acetone
was evaporated in vacuo. Brine was added to the residue, and the
mixture was extracted with tert-butyl methyl ether (4 × 5 mL).
The combined extracts were dried with MgSO4 and concentrated
in vacuo to give a colorless oil, which was shown by NMR
spectroscopy to contain cyclopent-3-en-1-yl mesylate (12) and
cyclopent-3-en-1-ol (14) in the ratio of 2:1, 1-butyl mesylate, and
a number of impurities but no product with a bicyclo[1.1.0]but-2-
ylcarbinyl group. The signals of 12 and 14 were identified by
comparison with those of authentic samples.
Solvolysis of 10 in [D4]Methanol/Sodium [D3]Methoxide. A
solution of NaOCD3 in DOCD3 [9 mg of sodium (0.4 mmol) had
been dissolved in DOCD3 (0.7 mL)] was added to the mesylate 10
(20 mg, 0.12 mmol), containing some 1-butyl mesylate, in an NMR
tube at 22 °C. After several minutes, a 1H NMR spectrum showed
the presence of 10, cyclopent-3-en-1-yl mesylate (12), and [D3]-
methyl cyclopent-3-en-1-yl ether (15) in the ratio of 20:1:2.4. This
ratio changed to 1.3:1:3.6 within 2.5 h, and 10 was consumed
completely after 8.5 h with the 12:15 ratio being 1:6.5. Simulta-
neously, 1-butyl mesylate was consumed to the extent of about 40%
and converted into 1-butyl [D3]methyl ether within 8.5 h. 1H NMR
of 15 (400 MHz; DOCD3; internal reference, DOCHD2 at δ )
3.31): δ 2.33 (apparent ddm, line distances 16, 3 Hz, 2H), 2.56
(apparent ddm, line distances 16, 7 Hz, 2H), 4.12 (tt, J ) 7.0, 3.2
Hz, 1H), 5.66 (m, 2H). 13C NMR of 15 (101 MHz; DOCD3; internal
reference, DOCD3 at δ ) 49.0): δ 41.1, 59.7 (m, JC,D ) 22 Hz),
82.1, 129.2.
Solvolysis of 11 in [D4]Methanol/Sodium [D3]Methoxide. A
stock solution of NaOCD3 in DOCD3 was prepared by dissolution
of sodium (42 mg, 1.83 mmol) in DOCD3 (3 mL). When a mixture
of 11 and 1-butyl tosylate was added to 0.7 mL of this solution or
one that had been diluted 1:1 with pure DOCD3, the substrates were
transformed to exo-bicyclo[1.1.0]but-2-ylcarbinyl [D3]methyl ether
(16) and 1-butyl [D3]methyl ether, respectively, within several days
at 22 °C. The progress of the reactions was monitored by NMR
spectroscopy. For the quantitative analysis, the integral of the signals
of all aromatic protons of the sample was taken as the internal
standard. The ratio of 11 and 1-butyl tosylate remained constant
throughout the reactions, and the conversion of 1-butyl tosylate to
1-butyl [D3]methyl ether was virtually quantitative.
On reaction of 13.5 mg (0.057 mmol) of 11 and 6.5 mg (0.028
mmol) of 1-butyl tosylate in 0.7 mL of the above stock solution
(0.43 mmol of NaOCD3 in DOCD3), 11 was consumed to the extent
of 37, 61, 83, 88, and 89% within 6, 9, 21.5, 26, and 30 h,
respectively, and the yield of 16 amounted to 30, 48, 67, 69, and
74% after the same time periods.
1
further use. H NMR (400 MHz, CDCl3): δ 1.44 (m, 1H), 1.71
(td, J ) 3.2, 1.4 Hz, 2H), 1.88 (qd, J ) 3.2, 2.2 Hz, 1H), 2.73
(tqd, J ) 7.3, 3.3, 0.5 Hz, 1H), 3.03 (s, 3H), 4.05 (d, J ) 7.3 Hz,
2H). 13C NMR (101 MHz, CDCl3): δ 2.0 (dsext, JC,H ) 205.8, 3.5
Hz, 2C), 29.8 (dddt, JC,H ) 170.5, 153.5, 12.2, 4.3 Hz, 1C), 37.9
(q, JC,H ) 139.0, 1C), 44.8 (dm, JC,H ≈ 156 Hz, 1C), 65.1 (t, JC,H
) 150.5, 1C).
In a preceding experiment, CDCl3 was used as the solvent and
the NMR spectra were taken at -30 °C without workup, which
clearly indicated the presence of 10. As compared to the data of
the salt-free solution obtained at 27 °C, the lower temperature and
the presence of HNEt3+Cl- caused upfield shifts of about 0.3 ppm
1
in the H NMR spectrum and of up to 1.0 ppm in the 13C NMR
spectrum. After the sample had been kept at room temperature for
a day, the spectra showed that 40% of the amount of 10 had been
converted into cyclopent-3-en-1-yl mesylate (12) and a second
4-substituted cyclopentene, which could not be identified due to
the signals of the impurities, whereas 12 was characterized by
comparison of its signals with those of an authentic sample. The
consumption of 10 was complete after 9 days at room temperature.
exo-Bicyclo[1.1.0]but-2-ylcarbinyl Tosylate (11). Sodium hy-
dride (50 mg, 2.1 mmol) was added cautiously in several portions
to a stirred solution of exo-9 (57 mg, 0.68 mmol), containing 20%
of 1-butanol, in anhydrous THF (3 mL), which was kept at 0-5
°C under nitrogen, within 30 min. Sodium hydride had been
purchased as a suspension in paraffin oil and washed carefully with
petroleum ether (bp 30-50 °C) before use. The mixture was stirred
overnight at room temperature, then cooled to -17 °C, and treated
dropwise with a solution of p-toluenesulfonyl chloride (190 mg,
1.0 mmol) in anhydrous THF (1 mL). Stirring was continued for 2
h at -15 °C. Then water (3 mL) was added, and the mixture was
extracted with diethyl ether (2 × 10 mL). The combined organic
phases were washed with 5% aqueous NaHCO3 (2 × 10 mL) and
On reaction of 6 mg (0.025 mmol) of 11 and 3 mg (0.013 mmol)
of 1-butyl tosylate in a mixture of 0.35 mL of the above stock
solution (0.22 mmol of NaOCD3 in DOCD3) and 0.35 mL of pure
DOCD3, 11 was consumed to the extent of 22, 66, 87, and 91%
within 4, 24, 48, and 62.5 h, respectively, and the yield of 16
amounted to 16, 47, 67, and 82% after the same time periods.
J. Org. Chem, Vol. 71, No. 3, 2006 1025

Bentley et al.
1H NMR of 16 (400 MHz; DOCD3; internal reference, DOCHD2
at δ ) 3.31): δ 0.51 (m, 1H), 1.02 (br t, J ) 6.1 Hz, 1H), 1.46-
1.49 (m, 3H), 3.31 (d, J ) 6.1 Hz, 2H). 13C NMR of 16 (101 MHz;
DOCD3; internal reference, DOCD3 at δ ) 49.0): δ 0.1, 28.5, 43.4,
73.1; the septuplet of the CD3 group was not observed due to its
low intensity.
added a solution of 2.5 equivalents (7 µL) of triethylamine (stored
over KOH) in dichloromethane (80 µL, AR grade). The solution
was cooled to -10 °C, and methanesulfonyl chloride (1.4 µL, <1
equiv) in dichloromethane (20 µL) was added in portions slowly
with stirring over 5-10 min. After stirring for another 5-10 min
at -10 °C, an aliquot of the solution was injected directly into AR
grade methanol (previously degassed by sonication) and further
sonicated to disrupt any aggregates of dichloromethane. The solution
was then transferred to a conductivity cell and thermostated, and
the change in conductance was monitored.
HPLC measurements were obtained from a solution of exo-
tosylate (11, 1-2 mg) in acetonitrile (40 µL) that was added to the
solvolysis medium (10 mL); 10 × 1 mL aliquots were then sealed
in glass ampules.31 Aliquots (20 µL) of quenched reaction mixtures
were analyzed using a Waters Novapak C18 column (15 cm) eluted
with 65% (v/v) methanol/water, with detection at 225 nm (A )
0.2). Rate constants were calculated from the peak areas of the
esters, and the “theoretical infinity” of the zero ester area was
assumed.
Solvolysis of 11 in a Mixture of 90% [D6]Acetone/Water and
Triethylamine. In an NMR tube, 20 mg of a 4:1 mixture of 11
(0.068 mmol) and 1-butyl tosylate (0.017 mmol) and triethylamine
(17 mg, 0.17 mmol) were dissolved in 90% (v/v) (CD3)2CO/H2O
(0.7 mL). A 1H NMR spectrum was taken right away. The sample
was then kept at room temperature, and NMR spectra of it were
taken over a period of 118 days. For the quantitative analysis, the
integral of the signals of all aromatic protons of the sample was
used as the internal standard. The products were (exo-bicyclo[1.1.0]-
but-2-ylcarbinyl)triethylammonium tosylate (17) from 11 and
1-butyltriethylammonium tosylate from 1-butyl tosylate. Within 2,
7, 14, 32, 56, and 118 days, 11 was consumed to the extent of 3,
11, 19, 34, 49, and 71%, respectively, whereas the yield of 17 was
determined to be 2, 5, 11, 22, 34, and 41% after the same time
periods. The ratio of unreacted 11 and 1-butyl tosylate remained
constant over 118 days, and the yield of 1-butyltriethylammonium
tosylate was about 80%.
Acknowledgment. This work is dedicated to Professor
Siegfried Hu¨nig on the occasion of his 85th birthday. It was
supported by the Deutsche Forschungsgemeinschaft (project:
EN 197/10). E.B. thanks the Deutscher Akademischer Aus-
tauschdienst for a fellowship. We are indebted to Prof. G.
Szeimies for drawing our attention to ref 6 and to Prof. W. Adam
and Mr. J. Bialas for gifts of dimethyldioxirane (ref 8).
1H NMR of 17 [400 MHz; (CD3)2CO/H2O, 9:1; internal
reference, CHD2COCD3 at δ ) 2.09]: δ 0.59 (br quint, J ) 1.1
Hz, 1H), 1.08 (br t, J ) 6.3 Hz, 1H), 1.37 (tt, JH,H ) 7.2, JH,N
)
1.7 Hz, 9H), 1.53 (td, J ) 2.9, 0.7 Hz, 1H), 1.88 (dt, J ) 2.9, 0.8
Hz, 2H), 2.34 (s, 3H), 3.39 (d, J ) 6.3 Hz, 2H), 3.48 (q, J ) 7.2
Hz, 6H), 7.18 (m, 2H), 7.71 (m, 2H). 13C NMR of 17 [101 MHz;
(CD3)2CO/H2O, 9:1; internal reference, (CD3)2CO at δ ) 29.9]: δ
Supporting Information Available: Absolute energies and
Cartesian coordinates of the structures of the cations 13, 18, and
25-28 as well as the 13C NMR spectra of the new compounds and
the solvolysis products. This material is available free of charge
via the Internet at http://pubs.acs.org.
1.2, 7.8, 21.2, 28.3, 35.0, 53.9 (t, 1JC,N ) 2.7 Hz), 57.6 (t, 1JC,N
)
3.3 Hz), 126.8, 129.1, 139.8, 144.9.
Kinetic Methods. Standard procedures, as described elsewhere,4
were used for conductometric measurements, but the following
slightly revised method was used for mesylates. To the alcohol
endo-9 (1.7 mg), in a Wheaton bottle fitted with a magnetic stirrer
and sealed with a septum (having a pressure release needle), was
JO0519918
(31) Bentley, T. W.; Gream, G. E. J. Org. Chem. 1985, 50, 1776-1778.
1026 J. Org. Chem., Vol. 71, No. 3, 2006