The effect of strain on reactivity: Poor leaving groups increase strain-induced inhibition of alkene-forming elimination

Article abstract of DOI:10.1039/a807506i

Elimination to form a carbon-carbon double bond exocyclic to a cyclopropane ring is inhibited by factors which increase from 1.4 to 10^4.5 as the leaving group becomes poorer; strain induced in the transition structure can amount to some 50% of

Full text of DOI:10.1039/a807506i

The effect of strain on reactivity: poor leaving groups increase strain-induced
inhibition of alkene-forming elimination
Luca Volta and Charles J. M. Stirling*
Department of Chemistry, The University, Sheffield, UK S3 7HF. E-mail: c.stirling@sheffield.ac.uk
Received (in Liverpool, UK) 28th September 1998, Accepted 6th October 1998
Elimination to form a carbon–carbon double bond exocyclic
to a cyclopropane ring is inhibited by factors which increase
from 1.4 to 104.5 as the leaving group becomes poorer; strain
induced in the transition structure can amount to some 50%
of the enthalpy difference between strained and unstrained
products.
cyclopropane and methylenecyclopropane9 although its origin
is under discussion.10
The open-chain halides 1 were obtained by homolytic
addition of sulfonyl halides to 2-methylpropene.11 Addition of
thiophenol to the alkenyl sulfone 2 gave sulfide 3 which on
oxidation gave the bis-sulfone 4 The cyclopropanes were
obtained by the routes of Scheme 1. Reactions were run in
ethanolic sodium ethoxide to allow direct comparisons with
earlier results; the product from the open-chain substrates was
the conjugated alkene 2 and non-conjugated alkene 2a,12 while
the cyclopropanes gave the ethoxy adduct 10 from slow
elimination followed by rapid addition to the electrophilic
methylenecyclopropane 7. The unlikely alternative course of
direct substitution was ruled out by the piperidine test in which,
for example, the bis-sulfone 9 failed to react with piperidine in
ethanol (too weakly basic) but reacted rapidly with piperidine in
ethanolic sodium ethoxide to give the piperidino derivative 11
(piperidine more nucleophilic than ethoxide).
The effect of strain on reactivity is a very familiar phenomenon,
but one which has only rarely been quantified.1 In order to
quantify the effects of strain on reactivity properly, systems of
known and defined strain energy are required (not always a
simple matter) and reactions of unambiguous mechanism are
also needed (not always a simple matter either). In this
connection, earlier work from these laboratories has examined
acceleration of 1,2-elimination by incorporation of the leaving
group in a strained ring.2–4 In such cases acceleration results
from the release of strain, effectively lowering the energy of the
transition structure relative to the starting material. We have
also examined the effect of strain in the inhibition of higher
order eliminations leading to strained ring products in which the
effects of strain are remarkably variable.5,6 Both situations have
been examined theoretically7,8 and the correlation between
theory and practice is good.
Reactions were followed by UV spectroscopy for reactions in
which the product alkene was detectable and otherwise by GC.
Reactions were first order in substrate and first order in base.
Before comment on the impact of strain differentials can be
made, it is crucial to be certain of the mechanism of the
reactions in each case. Two methods to throw light on the
mechanisms have been adopted; the observed rate constants
have been compared with the rates of ionisation obtained by
interpolation on a Taft plot13 of kionisation in ethanolic sodium
ethoxide versus the inductive constant s*. It can be seen that for
the leaving groups SO2Ph, SPh and OMe, the rate constant for
ionisation is greater than the elimination rate constant. This
points for each case to the (E1cB)R mechanism, in which a pre-
equilibrium with the carbanion is established with the base–
solvent system. The rate-determining step in each case,
therefore, is the expulsion of the leaving group from the
intermediate carbanion. The second procedure was to carry out
We now report on the kinetics of formation of methylene-
cyclopropanes in activated 1,2-eliminations which reveal the
impact of strain as the transition structure changes. The systems
we have examined, involving five different leaving groups (Z)
are in Table 1, giving rate constants for the unstrained [eqn. (1)]
and strained reactions [eqn. (2)] respectively. Making the
SO2Ph
SO2Ph
(1)
Z
SO2Ph
Z
SO2Ph
2
reactions in EtOD and to examine by H NMR spectroscopy
(2)
recovered starting material for deuterium incorporation. In all of
these cases, starting material had exchanged considerably with
the solvent, confirming the conclusions from interpolation. By
contrast, with the halogen leaving groups, the observed rate
constants in all cases are close to or greater than the interpolated
ionisation rate constants and no incorporation of deuterium
occurred in exchange experiments. These observations point
assumption that strain energies are unaffected by substituents,
the strain energy difference between substrate and product for
the methylenecyclopropane systems of eqn. (2) amounts to 50
kJ mol21. This is the strain energy difference between
Table 1 Elimination to form unstrained and strained alkenes
Rate constants/mol21 dm3 s21
Open chain
Cyclopropane
a
a,b
a,b
kEtO2 a
kion.
krel unstrained:strained
2
Z
kEtO
kion.
Br
Cl
SO2Ph
SPh
OMe
2.3 3 102
7.8 3 101
3.44
2.3 3 101
4.1 3 101
8.8 3 102
4.8 3 1021
3.1
3.2 3 102
3.2 3 101
5.7 3 101
1.2 3 103
6.7 3 1021
5.3 3 1021
0.7
1.4
230
6000
29 000
5.5 3 101
1.5 3 1022
1.0 3 1025
1.5 3 1029 d
6.6 3 1022
4.3 3 1025 c
a For reactions in EtONa–EtOH at 25 °C. b See text. c See ref. 13; d By extrapolation from an Arrhenius plot.
Chem. Commun., 1998, 2481–2482
2481

PhSO2
(E1cB)R mechanism. Our observation of probable E2 and/or
(E1cB)I mechanisms for the halides mentioned above, suggests
higher nucleofugalities for the halogens, as would be ex-
pected.
i
Hal
1
The results of Table 1 show that as the nucleofugality of the
leaving group decreases, so the ratio of the reactivities of the
unstrained to the strained substrates increases. This reveals a
consistent picture in which as the nucleofugality of the leaving
group decreases, so the degree of double bond character in the
transition structure increases and the additional strain of the
double bond exocyclic to the cyclopropane ring is increasingly
felt.
ii
PhSO2
PhSO2
PhSO2
2a
2
iii
PhSO2
iv
This leaves the important question of the extent of strain
inhibition in these reactions. In the system with the largest
unstrained to strained reactivity ratio, i.e. with Z = OMe, the
inhibition amounts to a factor of some 29,000 or about 26 kJ
mol21 in DG‡. This amounts to about 50% of the strain energy
difference between strained and unstrained products. When the
leaving group is halogen, the unstrained and strained substrates
have almost identical reactivities and there appears to be so little
double bond character in the transition structure that reactions
are little inhibited by formation of a strained alkene product.
We thank the University of Sheffield for the support of this
work and Elaine Frary for preliminary experiments.
PhSO2
PhS
4
3
Hal
Hal
SPh
Hal
v
iii
5
iv
i
SO2Ph
SO2Ph
SO2Ph
Hal
iii
ii
SPh
Notes and references
1 C. J. M. Stirling, Tetrahedron, 1985, 41, 1613.
2 H. A. Earl and C. J. M. Stirling, J. Chem. Soc., Perkin Trans. 2, 1987,
1273.
3 D. J. Young and C. J. M. Stirling, J. Chem. Soc., Perkin Trans. 2, 1996,
425.
4 S. W. Roberts and C. J. M. Stirling, J. Chem. Soc., Chem. Commun.,
1991, 170
8
9
7
6
vi
iv
SO2Ph
SO2Ph
SO2Ph
SO2Ph
vii
N
OEt
10
11
5 F. Benedetti and C. J. M. Stirling, J. Chem. Soc., Perkin Trans. 2, 1986,
605.
6 S. M. Jeffery and C. J. M. Stirling, J. Chem. Soc., Perkin Trans. 2, 1993,
2163.
Scheme 1 Reagents and conditions: i, PhSO2Hal, AIBN, benzene, 90 °C, 72
h; ii, Et3N, PhMe; iii, PhSNa, EtOH; iv, H2O2, AcOH; v, Hal2, THF; vi,
EtONa, EtOH; vii, EtONa, EtOH, piperidine.
7 S. M. van der Kerk, J. W. Verhoeven and C. J. M. Stirling, J. Chem.
Soc., Perkin Trans. 2, 1985, 1355.
8 G. Tonachini, F. Bernardi, H. B. Schlegel and C. J. M. Stirling, J. Chem.
Soc., Perkin Trans. 2, 1988, 705.
9 J. F. Liebman and A. Greenberg, in The Chemistry of the Cyclopropyl
Group, ed. Z. Rappoport, Wiley, Chichester, ch. 18, 1987.
10 W. T. G. Johnson and W. T. Borden, J. Am. Chem. Soc., 1997, 119,
5930.
11 S. Caddick, C. L. Shering and S. N. Woolman, Chem. Commun., 1997,
171.
12 I. Sataty and C. Y. Myers, Tetrahedron Lett., 1974, 4161.
13 P. J. Thomas and C. J. M. Stirling, J. Chem. Soc., Perkin Trans. 2, 1977,
1909.
14 D. R. Marshall, P. J. Thomas and C. J. M. Stirling, J. Chem. Soc., Perkin
Trans. 2, 1977, 1898.
either to the E2 mechanism, in which departure of the leaving
group is concerted with b-proton removal, or to the (E1cB)I
mechanism, in which it is not. For the latter mechanism, a close
2
similarity between kEtO . and kionisation is to be expected. This is
true for the chlorides, but the comparison for the bromides
suggests concerted mechanisms.
For 1,2-eliminations, it can reasonably be concluded that the
more difficult the leaving group is to expel, the greater is the
degree of double-bond character in the transition structure
required to expel it. In earlier work14 we were able to compare
accurately the leaving abilities of a series of groups placed b to
a sulfonyl-stabilised carbanion. These groups included SO2Ph,
SPh and OMe in descending order of nucleofugality. Nucleo-
fugalities of halide leaving groups could not be assigned
because, as in the present work, the reactions did not follow the
Communication 8/07506I
2482
Chem Commun., 1998, 2481–2482