Mechanism of the diphenyldisulfone-catalyzed isomerization of alkenes. Origin of the chemoselectivity: Experimental and quantum chemistry studies

Article abstract of DOI:10.1021/ja0579210

Polysulfone- and diphenyldisulfone-catalyzed alkene isomerizations are much faster for 2-alkyl-1-alkenes than for linear, terminal alkenes. The mechanism of these reactions has been investigated experimentally for the isomerization of methylidenecyclopentane into 1-methylcyclopentene, and theoretically [CCSD(T)/6-311++G(d,p)//B3LYP/6-311++G(d,p) calculations] for the reactions of propene and 2-methylpropene with a methanesulfonyl radical, MeSO _2?. On heating, polysulfones and (PhSO₂)₂ equilibrate with sulfonyl radicals, RSO_2?. The latter abstract allylic hydrogen atoms in one-step processes giving allylic radical/RSO ₂H pairs that recombine within the solvent cage producing the corresponding isomerized alkene and RSO_2?. The sulfinic acid, RSO₂H, can diffuse out from the solvent cage (H/D exchange with MeOD,D₂O) and reduce an allyl radical. Calculations did not support other possible mechanisms such as hydrogen exchange between alkenes, electron transfer, or addition/elimination process. Kinetic deuterium isotopic effects measured for the (PhSO₂)₂-catalyzed isomerization of methylidenecyclopentane and deuterated analogues and calculated for the H abstraction from 2-methylpropene and deuterated analogues by CH ₃-SO_2? are consistent also with the one-step hydrogen transfer mechanism. The high chemoselectivity for this reaction is not governed by an exothermicity difference but by a difference in ionization energies of the alkenes. Calculations for CH₃SO_2? + propene and CH₃SO_2? + 2-methylpropene show a charge transfer of 0.34 and 0.38 electron, respectively, from the alkenes to the sulfonyl radical in the transition states of these hydrogen abstractions.

Full text of DOI:10.1021/ja0579210

Published on Web 05/24/2006
Mechanism of the Diphenyldisulfone-Catalyzed Isomerization
of Alkenes. Origin of the Chemoselectivity: Experimental and
Quantum Chemistry Studies
Dean Markovic´,‡ Adria´n Varela-AÄ lvarez,† Jose´ Angel Sordo,*,† and Pierre Vogel*,‡
Contribution from the Laboratorio de Qu´ımica Computacional, Departamento de Qu´ımica F´ısica
y Anal´ıtica, UniVersidad de OViedo, OViedo, Principado de Asturias, Spain, and Laboratoire de
glycochimie et de synthe`se asyme´trique, Ecole Polytechnique Fe´de´rale de Lausanne (EPFL),
CH-1015 Lausanne-Dorigny, Switzerland
Received December 5, 2005; Revised Manuscript Received April 18, 2006; E-mail: pierre.vogel@epfl.ch; jasg@uniovi.es
Abstract: Polysulfone- and diphenyldisulfone-catalyzed alkene isomerizations are much faster for 2-alkyl-
1-alkenes than for linear, terminal alkenes. The mechanism of these reactions has been investigated
experimentally for the isomerization of methylidenecyclopentane into 1-methylcyclopentene, and theoretically
[CCSD(T)/6-311++G(d,p)//B3LYP/6-311++G(d,p) calculations] for the reactions of propene and 2-meth-
•
ylpropene with a methanesulfonyl radical, MeSO2 . On heating, polysulfones and (PhSO2)2 equilibrate with
•
sulfonyl radicals, RSO2 . The latter abstract allylic hydrogen atoms in one-step processes giving allylic
radical/RSO2H pairs that recombine within the solvent cage producing the corresponding isomerized alkene
•
and RSO2 . The sulfinic acid, RSO2H, can diffuse out from the solvent cage (H/D exchange with MeOD,D2O)
and reduce an allyl radical. Calculations did not support other possible mechanisms such as hydrogen
exchange between alkenes, electron transfer, or addition/elimination process. Kinetic deuterium isotopic
effects measured for the (PhSO2)2-catalyzed isomerization of methylidenecyclopentane and deuterated
analogues and calculated for the H abstraction from 2-methylpropene and deuterated analogues by CH3-
•
SO2 are consistent also with the one-step hydrogen transfer mechanism. The high chemoselectivity for
this reaction is not governed by an exothermicity difference but by a difference in ionization energies of the
alkenes. Calculations for CH3SO2• + propene and CH3SO2• + 2-methylpropene show a charge transfer of
0.34 and 0.38 electron, respectively, from the alkenes to the sulfonyl radical in the transition states of
these hydrogen abstractions.
Introduction
from various olefins, isomerizations 1 f 4 occurred in the
absence of SO2.3 The ESR spectrum of the polysulfone showed
typical signals for carbon-centered and sulfonyl radicals.4 Alkene
isomerizations 1 f 4 were also induced upon UV irradiation
or heating (80-120 °C) of 1 in the presence of a catalytic
amount of (PhSO2)2 (diphenyldisulfone).5 Both the polysulfone-
and diphenyldisulfone-catalyzed alkene isomerizations followed
zeroth-order rate laws and were inhibited by radical scavenging
agents. Importantly, these reactions were much better yielded
than the SO2-induced isomerizations, the latter being ac-
Double bond migration in alkenes can be induced by SO2.
The process is generally explained by invoking an ene reaction
between an SO2 molecule and the alkene 1 to give a â,γ-
unsaturated sulfinic acid intermediate 2 which then undergoes
a [1,3]-sigmatropic shift to form an isomeric â,γ-unsaturated
sulfinic acid 3 followed by a retro-ene reaction that eliminates
SO2 (Scheme 1) and produces the isomeric alkene 4.1
In the case of methylidenecyclopentane and related alkenes
we found that their isomerization was inhibited by radical-
scavenging agents such as TEMPO (2,2,6,6-tetramethyl-1-
piperidinyloxy free radical) and Bu3SnH. Kinetic measurements
showed an induction period followed by a zeroth-order reaction.
During the induction period a white precipitate of a 1:1
copolymer of the alkene 1 and SO2, a polysulfone, was formed.2
When pure alkenes 1 were exposed to pure polysulfones arising
(2) Early work on the copolymerization of alkenes on SO2, see e.g.: (a) Ryden,
L. L.; Glavis, F. J.; Marvel, C. S. J. Am. Chem. Soc. 1937, 59, 1014-
1015. (b) Marvel, C. S.; Glavis, F. J. J. Am. Chem. Soc. 1938, 60, 2622-
2626. (c) Marvel, C. S.; Sharkey, W. H. J. Am. Chem. Soc. 1939, 61, 1603.
(d) Kharasch, R. S.; Sternfeld, E. J. Am. Chem. Soc. 1940, 62, 2559-
2560. (d) Snow, R. D.; Frey, F. E. J. Am. Chem. Soc. 1943, 65, 2417-
2418.
(3) Markovic´, D.; Vogel, P. Angew. Chem., Int. Ed. 2004, 43, 2928-2930.
(4) (a) Zutty, N. L.; Wilson, C. W., III. Tetrahedron Lett. 1963, 2181-2188.
(b) Flockhart, B. D.; Ivin, K. J.; Pink, R. C.; Sharma, B. C. J. Chem. Soc.
D 1971, 339-340. (c) Flockhart, B. D.; Ivin, K. J.; Pink, R. C.; Sharma,
B. C. In Nobel Symposium 22; Kinel, P.-O., Ed.; Almqvist and Wiksell:
Stockholm, Sweden, 1973; pp 17-21. (d) Kearney, J. J.; Clark, H. G.;
Stannet, V.; Campbell, D. J. Polym. Sci., Polym. Chem. Ed. 1971, 9, 1197-
1202. (e) Chambers, S. A.; Fawcett, A. H.; Fee, S.; Malone, J. F.; Smyth,
U. Macromolecules 1990, 23, 2757-2765.
(5) (a) Kice, J. L.; Pawlowski, N. E. J. Am. Chem. Soc. 1964, 86, 4898-4904.
(b) Kobayashi, M.; Tanaka, K.; Minato, H. Bull. Chem. Soc. Jpn. 1972,
45, 2906-2909. (c) Ho Huu, T.; Ito, O.; Iino, M.; Matsuda, M. J. Phys.
Chem. 1978, 82, 314-319.
† Universidad de Oviedo.
‡ Ecole Polytechnique Fe´de´rale de Lausanne.
(1) (a) Masilamani, D.; Rogic, M. M. J. Am. Chem. Soc. 1977, 99, 5219-
5220. (b) Masilamani, D.; Rogic, M. M. J. Am. Chem. Soc. 1978, 100,
4634-4635. (c) Raasch, M. S.; Smart, B. E. J. Am. Chem. Soc. 1979, 101,
7733. (d) Masilamani, D.; Reumann, M. E.; Rogic, M. M. J. Org. Chem.
1980, 45, 4602-4605. (e) Hiscock, S. D.; Isaacs, N. S.; King, M. D.; Sue,
R. E.; White, R. H.; Young, D. J. J. Org. Chem. 1995, 60, 7166-7169. (f)
Braveman, S. In The Chemistry of Sulfinic Acids, Esters, and Their
DeriVatiVes; Patai, S., Ed.; Wiley: New York, 1990; pp 298-300. (g)
Baudin, J.-B.; Julia, S. Bull. Soc. Chim. Fr. 1995, 132, 196-214.
9
7782
J. AM. CHEM. SOC. 2006, 128, 7782-7795
10.1021/ja0579210 CCC: $33.50 © 2006 American Chemical Society

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
Scheme 1
sulfonyl radicals, in particular PhSO2 , can combine9 into a
sulfinyl-sulfonyl anhydride, PhS(O)-OSO2Ph,10 a trace of
water could generate sulfinic and sulfonic acids (RSO2H,
RSO3H) capable of catalyzing the alkene isomerization by
protonation and deprotonation or acid addition and elimination
(mechanism D, Scheme 5).11
•
Scheme 2. Selectivity of SO2 and Polysulfone-Catalyzed Alkene
Isomerizations
The benzenesulfonyl radical engendered by thermal decom-
position of diphenyldisulfone is capable of abstracting a
hydrogen atom from diphenylmethane as demonstrated by the
formation of 1,1,2,2-tetraphenylethane.12 Sulfonyl radicals are
electrophilic13 and able to be added to alkenes, a key reaction
during polymerization of alkenes promoted by sulfonyl radicals
and during the copolymerization of alkenes and sulfur dioxide.2
In mechanism E of Scheme 5, we envision that the formation
of allyl radicals (e.g., 6) and sulfinic acids (e.g., PhSO2H) is
not a one-step hydrogen atom transfer as involved in mechanism
A and C of Scheme 4 but a two-step process involving first the
companied by polymer formation.3 Furthermore, the polysul-
fone- and diphenyldisulfone-catalyzed alkene isomerization were
chemoselective in the sense that linear, terminal alkenes (1-
substituted ethylenes) and 1,2-dialkylethylenes were not isomer-
ized at all, whereas 2-alkyl substituted alk-1-enes were isomer-
ized (Scheme 2). This discovery led us to invent new strategies
for the protection and deprotection of alcohols and the semi-
protection of polyols. Methallyl, prenyl, and methylprenyl ethers
undergo selective cleavages catalyzed either by diphenyldisul-
fone6 or by the solid polysulfone PS (generated from the
copolymerization of SO2 with methylidenecyclopentane) under
neutral conditions (Scheme 3). Allyl ethers and other alcohol
protective groups were not cleaved under these conditions.7,8
In a preliminary communication3 we proposed that the PS- and
(PhSO2)2-catalyzed isomerization of methylidenecyclopentane
(5) imply the generation of allyl radical intermediate 6 which
can diffuse out of the polymer or from the solvent cage and
react with another molecule of 5 to generate product 7
(1-methylcyclopentene) and another allyl radical 6 (chain
process, mechanism A + B of Scheme 4). This hypothesis was
consistent with the observation that a 1:1.06 mixture of 5 and
hexadeuterated derivative 8 led, in the presence of PS, to a
mixture of 1-methylcyclopentenes with CH3, CH2D, CHD2, and
CD3 groups.
•
addition of sulfonyl radical RSO2 onto the alkene with
generation of the alkyl radicals of types 10 and 11. In a second
step, which can be fast because of the radical-assisted heteroly-
sis14 elimination of RSO2H and formation of the allyl radical,
intermediate 12 occurs.15 This mechanism would explain the
chemoselectivity of our alkene isomerizations as secondary
radicals 10 would form less readily than the corresponding
tertiary alkyl radicals 11.
Alternatively, and as found for polar radicals,16,17 the
hydrogen atom transfer is a two-step process starting by an
electron transfer to generate radical cations of type 13 and 14,
followed by proton transfer to the sulfinate anion (mechanism
F, Scheme 5). This mechanism would also explain the chemose-
lectivity of our alkene isomerizations. Quantum calculations,
kinetic deuterium isotopic effects, and other experiments
presented in this work favor mechanism A + C; i.e., they
support one-step hydrogen transfer from the allylic C-H of
alkene to the sulfonyl radical. The hydrogen/deuterium scram-
We present here new experiments that refute this hypothesis
and shall show that the alternative mechanism A + C of Scheme
4 interprets the data better. Instead of direct hydrogen (deute-
rium) atom transfer from 5 (8) to the allyl radical 6 (9), the
hydrogen (deuterium) atom is transferred via the sulfinic acid,
RSO2H(D), generated by the reaction of the alkene with sulfonyl
radicals. It is the intermediate allyl radicals 6 (9) that can diffuse
away from the polysulfone (or the solvent cage in the case of
(PhSO2)2-catalyzed isomerization) and encounter another sulfinic
acid moiety (mechanism C of Scheme 4). As it is known that
(9) (a) Da Silva Correa, C. M. M.; Lindsay, A. S.; Waters, W. A. J. Chem.
Soc. C 1968, 6, 1872-1874. (b) Bennett, J. E.; Brunton, G.; Gilbert, B.
C.; Whittall, P. E. J. Chem. Soc., Perkin Trans. 2 1988, 1359-1364. (c)
Freeman, F.; Keindl, M. C. Sulfur Reports 1985, 4, 231-305.
(10) Boehme, H.; Meyer-Dulheuer, K. H. Liebigs Ann. Chem. 1965, 688, 78-
93.
(11) This mechanism is also compatible with inhibition by radical scavenging
agents; the latter inhibit the combination of two sulfinyl radicals.
(12) Kice, J. L.; Gabrielsen, R. S. J. Org. Chem. 1970, 35, 1010-1015.
(13) (a) Hazell, J. E.; Ivin, K. J. Trans. Faraday Soc. 1962, 58, 176-185. (b)
Lee, K. H. Tetrahedron 1970, 26, 1503-1507. (c) Da Silva Correa, C. C.
M.; Waters, W. A. J. Chem. Soc., Perkin Trans. 2 1972, 1575-1577. (d)
Takahara, Y. M.; Matsuda, M. Bull. Chem. Soc. Jpn. 1976, 49, 2268-
2271. (e) Gozdz, A. S.; Maslak, P. J. Org. Chem. 1991, 56, 2179-2189.
(f) Bertrand, M. P. Organic Preparations and Procedures International
1994, 26, 257-290. (g) Chatgilialoglu, C. In The Chemistry of Sulphones
and Sulphoxides; Patai, S., Rappoport, Z., Stirling, C. J. M., Eds.; John
Willey & Sons: Ltd. Chichester, 1988; Chapter 25, pp 1089-1113. (h)
Wang, S.-F.; Chuang, C.-P.; Lee, W.-H. Tetrahedron 1999, 55, 6109-
6118.
(6) Markovic´, D.; Vogel, P. Org. Lett. 2004, 6, 2693-2696.
(7) Markovic´, D.; Steunenberg, P.; Ekstrand, M.; Vogel, P. Chem. Commun.
2004, 2444-2445.
(8) For reviews on protection and deprotection of alcohols, see: (a) Kocien˜ski,
P. J. In Protecting Groups; Enders, D., Noyori, R., Trost, B. M., Eds.;
Thieme Verlag: Stuttgart, 1994. (b) Jarowicki, K.; Kocien˜ski, P. J. J. Chem.
Soc., Perkin Trans. 1 1999, 1589-1616. (c) Guibe´, F. Tetrahedron 1997,
53, 13509-13556. (d) Guibe´, F. Tetrahedron 1998, 54, 2967-3042. (d)
Stirling, C. J. M. In The Chemistry of Sulphinic Acids, Esters, and their
DeriVatiVes; Patai, S., Ed.; Wiley: Chichester, 1990; pp 1-7.
(14) (a) Cozens, F. L.; O’Neill, M.; Bogdanova, R.; Schepp, N. J. Am. Chem.
Soc. 1997, 119, 10652-10659 and references cited therein. (b) Zipse, H.
Acc. Chem. Res. 1999, 32, 571-578. (c) Horner, J. H.; Bagnol, L.;
Newcomb, M. J. Am. Chem. Soc. 2004, 126, 14979-14987.
(15) The elimination of RSO2H from the alkyl radical might be concerted or
-
become a radical-cation/RSO2 pair.
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7783

A R T I C L E S
Markovic´ et al.
Scheme 3. Chemoselective Cleavage of Alkyl Substituted Allyl Ethers under Neutral Conditions and without Metal Catalyst
Scheme 4. Possible Mechanisms for the Scrambling of H/D
Scheme 5. Other Possible Mechanisms for the Alkene Isomerizations Catalyzed by Polysulfones and Diphenyldisulfone
Theoretical Methods
bling presented in Scheme 4 is not due to allyl radical/alkene
reactions but to diffusion of the allyl radical and/or sulfinic acid
intermediates out of the solvent cage, a process that competes
with the transfer of a hydrogen atom from it to the allyl radical
intermediate within the initial solvent cage.
Propene, 2-methylpropene, and a methanesulfonyl radical were
chosen as model systems to carry out a theoretical mechanistic analysis
on the sulfonyl radical-catalyzed alkene isomerization.
9
7784 J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
Geometry optimizations were carried out using Density Functional
Theory (DFT)18 with the Becke three-parameter Lee-Yang-Parr19,20
(B3LYP) hybrid functional and Pople’s valence triple-ú 6-311++G-
(d,p) basis set.21 All the geometrical parameters were fully optimized,
and all the structures located on the potential energy surfaces (PESs)
were characterized as minima or transition structures by computing
the corresponding Hessian matrices and examining the number of
imaginary frequencies from them.
The energy predictions were improved in the case of the abstraction
reactions of propene and 2-methylpropene with the methanesulfonyl
radical by performing single-point CCSD(T)/6-311++G(d,p)//B3LYP/
6-311++G(d,p) calculations. As shown below, the chemoselectivity
observed experimentally for these reactions (no isomerization was
detected when using 1-substituted ethylenes and 1,2-dialkylethylenes)
will arise from the comparison between the substantially similar energy
profiles for the propene and 2-methylpropene reactions. Therefore, a
high level of accuracy is required to ensure the validity of the
conclusions.
Graphical analyses of the imaginary frequencies of the transition
structures as well as intrinsic reaction coordinate (IRC) calculations in
the mass-weighted internal coordinates22 allowed us to interconnect the
different structures located on the PESs and then construct the
corresponding energy profiles.
Solvation effects were estimated by using a continuum solvent
description: the so-called polarized continuum model (PCM).23 Cal-
culations were carried out for dichloromethane with a dielectric constant,
ꢀ, of 8.93. The atomic radii were taken from the universal force field
(UFF).24 The ion convention standard state (1M)25 was used to compute
the changes in Gibbs free energies (∆G).
The Gaussian98 and Gaussian03 packages of programs27 were used
to carry out all these calculations. Isotope effects were computed by
means of the QUIVER program,28 which employs the Bigeleisen-
Mayer formulation29 within the transition-state-theory approximation.
Experimental Results and Discussion
A. Mechanism of the Hydrogen Exchange between Alk-
enes. Isomerization of methylidenecyclopentane (5) into 1-me-
thylcyclopentene (7) catalyzed by the solid polysulfone PS
generated by copolymerization of 5 and SO2 occurs already at
0 °C. The rate of the isomerization does not depend on the mode
of preparation of PS and upon the amount of aqueous NaOH
used to neutralize the free SO2H groups it contains. The PS as
well as the diphenyldisulfone-catalyzed isomerizations 5 f 7
have similar rates (at a given temperature) in CH2Cl2 or heptane.
The PS-catalyzed isomerization 5 f 7 is strongly retarded in
polar solvents such acetone, EtOAc, or DMF. This is not the
case for the disulfone-catalyzed isomerization. We attribute this
different behavior between the PS and disulfone catalyst to the
swelling properties of the polar solid polysulfone catalyst. Polar
solvent diffuse with greater difficulty out of the polymer and
do not leave a chance for the alkene to get into the active site
of the solid catalyst.30 When THF/H2O mixtures were used as
solvent, the isomerization 5 f 7 catalyzed by (PhSO2)2 at 80
°C was not stopped on adding NaHCO3. The reaction was
slowed, however, due to decomposition of diphenyldisulfone
under those conditions. Finally, benzenesulfinic acid (10 mol
%) was not capable of inducing isomerization of meth-
ylidenecyclopentane (5) into 1-methylcyclopentene (7) during
3 h at room temperature. These observations demonstrate that
it cannot be neither protonation/deprotonation nor sulfinic acid
addition/elimination that is responsible for the alkene isomer-
ization. Thus mechanism D of Scheme 5 can be ruled out. These
results also demonstrate that the sulfinic acid intermediate
RSO2H, formed by abstraction of a hydrogen atom from the
alkene, does not have to diffuse out of the solvent cage (in the
case of the disulfone-catalyzed isomerization) to deliver an
The thermodynamic functions (∆H, ∆S, and ∆G) were estimated
within the ideal gas, rigid rotor, and harmonic oscillator approxima-
tions.26 A temperature of 298 K was assumed.
(16) (a) Karki, S. B.; Dinnocenzo, J. P.; Jones, J. P.; Korzekwa, K. R. J. Am.
Chem. Soc. 1995, 117, 3657-3664. (b) Manchester, J. I.; Dinnocenzo, J.
P.; Higgins, L. A.; Jones, J. P. J. Am. Chem. Soc. 1997, 119, 5069-5070.
(c) Shaffer, C. L.; Morton, M. D.; Hanzlik, R. P. J. Am. Chem. Soc. 2001,
123, 8502-8508. (d) Shaffer, C. L.; Harriman, S.; Koen, Y. M.; Hanzlik,
R. P. J. Am. Chem. Soc. 2002, 124, 8268-8274. (e) Sato, H.; Guengerich,
F. P. J. Am. Chem. Soc. 2000, 122, 8099-8100. (f) Guengerich, F. P.;
Yun, C.-H.; Macdonald, T. L. J. Biol. Chem. 1996, 271, 27321-27329.
(g) Migliavacca, E.; Carrupt, P.-A.; Testa, B. HelV. Chim. Acta 1997, 80,
1613-1626. (h) Goto, Y.; Watanabe, Y.; Fukuzumi, S.; Jones, J. P.;
Dinnocenzo, J. P. J. Am. Chem. Soc. 1998, 120, 10762-10763. (i) Wright,
J. S.; Johnson, E. D.; DiLabio, G. A. J. Am. Chem. Soc. 2001, 123, 1173-
1183. (j) Fukuzumi, S.; Shimoosako, K.; Suenobu, T.; Watanabe, Y. J.
Am. Chem. Soc. 2003, 125, 9074-9082. (k) Itoh, S.; Kumei, H.; Nagatomo,
S.; Kitagawa, T.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 2165-2175.
(l) Baciocchi, E.; Bietti, M.; Gerini, M. F.; Lanzalungo, O. J. Org. Chem.
2005, 70, 5144-5149. (m) Nakanishi, I.; Kawashima, T.; Ohkubo, K.;
Kanazawa, H.; Inami, K.; Mochizuki, M.; Fukuhara, K.; Okuda, H.; Ozawa,
T.; Itoh, S.; Fukuzumi, S.; Ikota, N. Org. Biomol. Chem. 2005, 3, 626-
629.
(27) (a) Frisch, M. J. et al. Gaussian98, revision A.7; Gaussian, Inc.: Pittsburgh,
PA, 1998. (b) Frisch, M. J. et al. Gaussian03, revision C.01; Gaussian,
Inc.: Wallingford, CT, 2004.
(28) Saunders, M.; Laidig, K. E.; Wolsberg, M. J. Am. Chem. Soc. 1989, 111,
8989-8994.
(29) Bigeleisen, J.; Mayer, M. G. J. Chem. Phys. 1947, 15, 261.
(30) One referee suggested that the mechanism of the (PhSO2)2-induced alkene
isomerization might not be the same as that of the polysulfone-induced
reaction. If sulfonyl-centered radicals should equilibrate with the polysulfone
already at room temperature, the C-S bond energy in sulfones of type
(dialkyl)C-SO2-alkyl is too high for such a homolysis. He calculates
(CBS-QB3) bond dissociation energies of 71.8 and 67.2 kcal/mol for the
C-S bonds in Me2SO2 and MeSO2-C(Me)2, respectively. Experimental
thermochemical data allow one to estimate to ca. 63 kcal/mol the C-S
bond energy of MeSO2C(Me)3 considering the C-S bond dissociation
energy in MeSO2Me of 73.5 kcal/mol31 and the difference in C-H bond
energies between Me-H and Me3C-H.31 The existence of the sulfonyl-
centered radical in our polysulfone PS derived from the copolymerization
of SO2 and methylidenecyclopentane is demonstrated by ESR.3 Polysulfones
are known to equilibrate with SO2 and alkenes on heating: the ceiling
temperature depends on the nature of the alkene used for the copolymer-
ization with SO2. It can be lower than 273 K with 2-alkylalk-1-enes,32 and
sulfonyl-centered radicals have been detected by ESR during the depo-
lymerization of polysulfones.33 During the copolymerization of alkenes and
SO2, sulfite moieties (CH2-S(O)-OC(R)2-) are formed concurrently with
sulfone moieties (CH2-SO2-C(R)2). The former undergo homolysis into
alkyl and sulfonyl radicals much more readily than the latter as the sulfones
are significantly more stable than their isomeric sulfinates.31b,c The sulfinate
impurities are the most probable source of the radicals formed in the initial
step of the polysulfone depolymerization process.34 With our solid
polysulfone PS Front-strain and Back-strain are expected to lower the
barriers of the C-S(sulfone) and C-O(sulfinate) homolyses, and highly
positive entropies of dissociation are expected for the C-O bond leakage
of sulfinates with high molecular mass.
(17) See also proton-coupled electron tranfer reactions, e.g.: (a) Cukier, R. I.;
Nocera, D. G. Annu. ReV. Phys. Chem. 1998, 49, 337-369. (b) Kohen,
A.; Klinman, J. P. Acc. Chem. Res. 1998, 31, 397-404. (c) Hammes-
Schiffer, S. Acc. Chem. Res. 2001, 34, 273-281. (d) Mayer, J. M.; Hrovat,
D. A.; Thomas, J. L.; Borden, W. T. J. Am. Chem. Soc. 2002, 124, 11142-
11147. (e) Olivella, S.; Anglada, J. M.; Sole´, A.; Bofill, J. M. Chem.s
Eur. J. 2004, 10, 3404-3410.
(18) Koch, W.; Holthausen, M. C. A Chemist’s Guide to Density Functional
Theory, 2nd ed.; Wiley-VCH: Weinheim, 2000, pp 308.
(19) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1992, 37, 785.
(20) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(21) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980,
72, 650-654.
(22) (a) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154-2161.
(b) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1990, 94, 5523-5527.
(23) (a) Cance`s, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032.
(b) Mennucci, B.; Cance`s, E.; Tomasi, J. J. Phys. Chem. B 1997, 101,
10506-10517. (c) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Chem.
Phys. 2001, 114, 5619-5701. (d) Cossi, M.; Scalmani, G.; Rega, N.;
Barone, V. J. Chem. Phys. 2003, 117, 43-54.
(24) Rappe´, A. K.; Casewit, C. J.; Colwell, K. S.; Godard, W. A., III; Shiff, W.
M. J. Am. Chem. Soc. 1992, 114, 10024.
(25) Cramer, C. J. Essentials of Computational Chemistry. Theories and Models;
Wiley: Chichester, 2002.
(26) McQuarrie, D. A. Statistical Thermodynamics; University Science Books:
Mill Valley, CA, 1973.
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7785

A R T I C L E S
Scheme 6
Markovic´ et al.
hydrogen atom to another allyl radical; the isomerization can
occur within the solvent cage or polymer active site.
To our surprise, when solutions of methylidenecyclopentane
(5) in C6D6 containing 5-50 mol % of azo compound 13 were
heated to 60-80 °C, only 15 was formed38 and 5 was recovered
unrearranged quantitatively (1H NMR analysis). These observa-
tions demonstrate that the hydrogen atom transfer from 5 to
the methallyl radical (14) is a very slow reaction that does not
compete with the dimerization of 14 into 15. This suggests that
the hydrogen atom transfer from 5 to the allyl radical 6 might
also be a very slow process that cannot be taken to be
responsible for isomerization 5 f 7 and for the hydrogen/
deuterium scrambling presented in Scheme 4. Mechanism A +
B cannot be retained, isomerization occurs, in part, within the
solvent cage of the allyl•RSO2H species, and the D/H scrambling
between 5 and 8 (Scheme 4) is due to competitive diffusion of
the allyl radical or/and the sulfinic acid species outside the
solvent cage (mechanism A + C of Scheme 4). This latter
hypothesis was confirmed by the following experiments.
On heating 5 (0.2 molar) in cyclohexane containing 4 to 10
equiv of MeOD in the presence of 10 mol % of (PhSO2)2,
isomerization 5 f 7 occurred and 7 was partially deuterated at
its methyl group. The proportion of 7:7-d1 (Scheme 7) varied
between 45:55 and 55:45. The incomplete monodeuteration,
although MeOD was used in excess, indicates that about 50%
of the isomerization 5 f 7 occurs within the solvent cage and
that about 50% occurs with concurrent diffusion of the allyl
radical or/and PhSO2H outside the solvent cage (assuming fast
deuterium hydrogen exchange once PhSO2H gets in contact with
MeOD). Similar results were obtained on treating 5 with 5 mol
% I2 in THF-d8/MeOD at 80 °C (Scheme 7). In this latter
experiment, the yield in 7 + 7-d1 was mediocre due to
concurrent formation of several other products. To confirm the
above hypothesis we ran an isomerization 5 f 7 in THF/D2O
and another one in THF-d8/D2O, both catalyzed by 10 mol %
diphenyldisulfone (80 °C). This generated pure 7 + 7-d1 in the
proportion 55:45 (10 to 30 equiv of D2O).39 On adding 20 mol
% of NaHCO3 to these solutions, the diphenyldisulfone-
On heating 5 (sealed tube, CD2Cl2) in the presence of AIBN
(2,2′-azobisisobutyronitrile, 10-50 mol %), no isomerization
5 f 7 could be observed after complete destruction of AIBN.
This shows that the 2-cyanoisopropyl radical is not capable of
abstracting a hydrogen atom from alkene 5, or if it can do it, it
does not lead to the isomerization 5 f 7. As reaction Me2-
(C(CN)• + 5 f 6 + Me2(CN)CH is endergonic31 we reasoned
that the 2-methylallyl radical would be a better initiator for the
isomerization 5 f 7 as reaction CH2dC(Me)CH2• + 5 f 6 +
isobutylene is estimated to be exergonic by -1 to -2 kcal/
mol.31 If this equilibrium should occur, it could lead to
isomerization 5 f 7 via the radical chain process 5 + 6 f 7 +
6 (mechanism A + B, Scheme 4). To test this hypothesis we
prepared 1,1′-azobis(2-methylpropene) (13) following a proce-
dure analogous to that reported for the preparation of 1,1′-azo-
(propene).35
N,N-Bis(ethoxycarbonyl)hydrazine (10)36 was reacted with
methallyl bromide in the presence of NaH in DMF giving 11
in 76% yield. Treatment of 11 with aqueous KOH at 100 °C
led to a poor yield (12%) in hydrazine 12. However, when
heated to 160 °C in an autoclave, 87% of 12 was obtained.
Known methods35 for the oxidation of 12 into the azo compound
13 such as treatment with HgO/HgSO4 in ether at 0 °C led to
the exclusive formation of 2,5-dimethylhexa-1,5-diene (15).
Finally we found that simple bubbling of air in a solution of 12
in cyclohexane at 0 °C led to its clean oxidation into 13 that
was isolated pure in 73% yield (Scheme 6). On heating 13 in
C6D6 at 80 °C, 2,5-dimethylhexa-1,5-diene (15) and N2 were
formed quantitatively, a reaction believed to imply the formation
of a 2-methylallyl radical intermediate (14).37
(31) (a) Lias, S. G.; Bartmess, J. E.; Liebman, D. F.; Holmes, J. L.; Levin, R.
D.; Mallard, W. G. J. Phys. Chem. Ref. Data Suppl. 1988, 17, 861 (http://
webbook. nist. gov.). (b) Liebman, J. F.; Crawford, K. S. K.; Slayden, S.
W. In Supplement S: The chemistry of sulphur-containing functional
groups; Patai, S., Rappoport, Z., Eds.; Wiley & Sons Ltd.: New York, 1993;
Chapter 4, pp 197-243. (c) Mackle, H. Tetrahedron 1963, 19, 1159-
1170. (d) Cox, J. D.; Pilcher, G. Thermochemistry of Organic and
Organometallic Compounds; Academic Press: New York, 1970.
(32) See e.g.: Cook, R. E.; Dainton, F. S.; Ivin, K. J. J. Polym. Sci. 1957, 26,
351-364.
(33) See e.g.: Zutty, N. L.; Wilson, C. W.; Potter, G. H.; Priest, D. C.;
Whitworth, C. J. J. Polym. Sci., Part A: Polym. Chem. 1965, 3, 2781-
2799 and ref 4.
(34) Zhao, Y.-L.; Jones, W. H.; Monnat, F.; Wudl, F.; Houk, K. N. Macro-
molecules 2005, 38, 10279-10285.
(35) Al-Sader, B. H.; Crawford, R. J. Can. J. Chem. 1970, 48, 2745-2754.
(36) Diels, O.; Borgwardt, E. Chem. Ber. 1920, 53, 150-158. Milcent, R.;
Guevrekian-Soghomoniantz, M.; Barbier, G. J. Heterocycl. Chem. 1986,
23, 1845-1848.
(37) (a) Callear, A. B.; Lee, H. K. Nature 1967, 213, 693-694. (b) Kochi, J.
K.; Krusic, P. J. J. Am. Chem. Soc. 1968, 90, 7157-7159. (c) Mitchell, T.
J.; Benson, S. W. Int. J. Chem. Kinet. 1993, 25, 931-955. (d) Getty, J. D.;
Liu, X.; Kelly, P. B. J. Chem. Phys. 1996, 104, 3176-3180 and references
cited therein.
(38) (a) Hefter, H. J.; Wu, C.-H. S.; Hammond, G. S. J. Am. Chem. Soc. 1973,
95, 851-855. (b) Roth, W. R.; Bauer, F.; Beitat, A.; Ebbrecht, T.;
Wu¨stefeld, M. Chem. Ber. 1991, 124, 1453-1460.
(39) Concurrent hydrolysis of diphenyldisulfone occurred under these conditions,
thus leading to incomplete isomerization of the alkene, unless more
diphenyldisulfone would be added to the reaction mixture. For the hydrolysis
of diphenyldisulfone, see: Kice, J. L.; Margolis, H. C.; Johnson, W. C.;
Wulff, C. A. J. Org. Chem. 1977, 43, 2933-2935.
9
7786 J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
Scheme 7. Deuterium Incorporation Implies Migration of the Allyl Radical or/and RSO2H outside the Solvent Cage of Allyl Radical +
RSO2H
•
Scheme 8. PhSO2 -Induced Isomerizations
transfer and subsequent proton transfer of Scheme 5)) we have
measured the kinetic deuterium isotopic effects defined in
Scheme 10. The results will be compared with related kinetic
isotopic effects calculated (see below) for model reactions. As
we shall see, quantum calculations as well as our experimental
results support mechanism A for the formation of allyl radical
intermediates.
Next the hexadeuterated methylidenecyclopentane 8 was
prepared along with the tetradeuterated derivative 19 and the
dideuterated derivative 20 (Scheme 10).42
catalyzed (80 °C) isomerization 5 f 7 led only to the formation
of nondeuterated isomerized alkene 7. By D-NMR analysis
The degree(s) of deuterium incorporation at C(2,5) and at
2
1
1
the olefinic center were measured by H NMR (residual H
signal compared with a 13C-H satellite) and by mass spec-
trometry. The rate constants for the diphenyldisulfone-catalyzed
isomerization (10 mol %) were measured by 1H NMR and 2D-
NMR in benzene-d6 at 80 °C for each compound 5, 8, 19, and
20 and for pairs of alkenes (1:1 mixtures of 5 + 8, 5 + 19, 5
+ 20, 19 + 20, 8 + 19, and 8 + 20). Internal reference was 10
mol % of toluene-d8; initial concentration of alkenes: 0.1-
0.15 M. The kinetic data so-obtained (all zeroth-order rate laws)
are shown in Table 1.
(toluene-d8 as internal reference), less than 3% of 7-d1 was
formed under the latter conditions.
To confirm that PhSO2• is indeed the species responsible for
the diphenyldisulfone-catalyzed alkene isomerization we gener-
ated it from other sources than (PhSO2)2. For instance, on
1
heating 5 in cyclohexane-d12 (sealed tube, H NMR analysis)
to 110 °C in the presence of 10 mol % of sulfones 1640 and
17,41 7 was formed quantitatively together with a few percents
of hexa-1,5-diene and 2,5-dimethylhexa-1,5-diene, respectively.
The above isomerizations 5 f 7 were inhibited on adding small
amounts (2-10 mol %) of (Bu3Sn)2 to the reaction mixture
(Scheme 8).
We notice that deuteration of 5 at C-2 and C-5 retards the
isomerization much more than deuteration at the terminal center
of the alkene.
We also verified that PhSO2H is capable of reducing allyl
radical intermediates by hydrogen transfer.1f Thus, on heating
the azo compound 13 with 2 equiv of PhSO2H in cyclohexane-
d12 at 40 °C (sealed tube, 1H NMR analysis), quantitative
formation of isobutylene (2-methylpropene) was observed
(Scheme 9).
Furthermore, we found that 4-bromoheptane can be reduced
smoothly into heptane on heating it to 120 °C in cyclohexane-
d12 in the presence of 1.3 equiv of PhSO2H and a 1 equiv of
(Bu3Sn)2.
If a single electron transfer should be the rate-determining
step, much smaller kH/k6D and kH/k4D values would be ob-
served.43 Similarly, if the rate determination step of the
isomerization 5 f 7 should be the addition of PhSO2• onto the
alkene to generate the tertiary radical 11 (R′ ) Me) (see Scheme
5) smaller kH/k6D and kH/k4D values would have been expected.
If mechanisms E or/and F of Scheme 5 should be followed for
the formation of the allyl radical 6 and PhSO2H, our kinetic
deuterium isotopic effects would imply that the second step of
these mechanisms would be rate-determining; i.e., either the
1,3-elimination 11 (R′ ) Me) f 6 + PhSO2H or the proton
transfer 14 (R′ ) Me) f 6 + PhSO2H is the slow step.43 When
the latter mechanism is followed (rate-determining proton
B. Mechanism of the Formation of Allyl Radical Inter-
mediates. To address the question of the mechanism of
formation of allyl radicals by reaction of sulfonyl radicals,
•
RSO2 , with alkenes (distinction between mechanisms A (direct
•
one-step hydrogen transfer), E (RSO2 radical addition and
(42) (a) Malloy, T. B., Jr.; Hedges, R. M.; Fisher, F. J. Org. Chem. 1970, 35,
4256-4257. (b) Malloy, T. B., Jr.; Fisher, F.; Laane, J.; Hedges, R. M. J.
Mol. Spectrosc. 1971, 40, 239-261. (c) Leriverend, M. L.; Leriverend, P.
Chem. Ber. 1976, 109, 3492-3495. (d) Wolkoff, P.; Holmes, J. L. Can. J.
Chem. 1979, 57, 348-354.
(43) (a) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic Molecules;
Wiley-Interscience: New York, 1984. (b) Carpenter, B. K. Determination
of Organic Reaction Mechanisms; Wiley-Interscience: New York, 1984.
(c) Koshino, N.; Saha, B.; Espenson, J. H. J. Org. Chem. 2003, 68, 9364-
9370. (d) Huynh, M. H. V.; Meyer, T. J. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 13138-13141. (e) Fokin, A. A.; Schreiner, P. R. Chem. ReV.
2002, 102, 1551-1593.
subsequent 1,3-elimination of RSO2H), and F (single electron
(40) (a) Van Zuydewijn, E. de R. Recl. TraV. Chim. Pays-Bas 1937, 56, 1047-
1062. (b) Backer, H. J.; Dost, N. Rec. TraV. Chim. Pays-Bas 1949, 68,
1143-1161.
(41) (a) Durst, T.; Tin, K. C.; Marcil, M. J. V. Can. J. Chem. 1973, 51, 1704-
1712. (b) Sataty, I.; Meyers, C. Y. Tetrahedron Lett. 1974, 4161-4164.
(c) Trost, B. M.; Schmuff, N. R.; Miller, M. J. J. Am. Chem. Soc. 1980,
102, 5979-5981. (d) Volta, L.; Stirling, C. J. M. J. Chem. Soc., Chem.
Commun. 1998, 2481-2482.
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7787

A R T I C L E S
Markovic´ et al.
Scheme 9. Reduction with PhSO2H
Scheme 10. Preparation of Deuterated Methylidenecyclopentanes (s ) deuteration degree at the marked positions)
Scheme 11
Table 1. Zeroth-Order Rate Constants (mol kg-1 min-1
benzene-d6, 80 °C, 10 mol % (PhSO2)2, 10 mol % toluene-d8) of
the Disulfone-Catalyzed Isomerizations of 5, 8, 19, 20 (Average of
3 Measurements at least)a
,
processes, and thus start a chain process (mechanism B)
responsible for the alkene isomerization. As we have found that
the 2-cyanoprop-2-yl radical and methallyl radical are not
capable of catalyzing the isomerization of methylidenecyclo-
pentane into 1-methylpentane; this pathway (mechanisms E +
G + B) can be ruled out. Indeed, as we shall see here below,
our quantum calculations predict a high barrier for the hydrogen
atom transfer between isobutylene and the methallyl radical,
and consequently such a pathway can also be discarded based
on theoretical considerations.
5 f 7-CH3
8 f 7-CD3
19 f 7-CH2D
20 f 7-CD2H
kH ) 0.001 43 ( 0.000 02
k6D ) 0.000 39 ( 0.000 01
k4D ) 0.000 48 ( 0.000 01
k2D ) 0.001 27 ( 0.000 03
Kinetic Deuterium Isotopic Effects (80 °C)
kH/k6D
kH/k4D
kH/k2D
k
2D/k4D
k2D/k6D
k4D/k6D
3.67 ( 4% 2.97 ( 3.5% 1.12 ( 3.8% 2.64 ( 4.5% 3.25 ( 5% 1.23 ( 4.8%
(4.21-4.57) (4.08-4.43) (1.02-1.03)
Theoretical Analysis of the Mechanism
a The theoretical estimates of the kinetic isotopic effects (in parentheses)
were obtained by employing the zero- and small-curvature tunneling models,
respectively (see the text for further details).
A. Mechanism of the Hydrogen Exchange between the
Alkene and Sulfonyl Radicals. Figures 1 and 2 contain the
B3LYP/6-311++G(d,p) geometries for the species involved in
the abstraction reactions between propene and 2-methylpropene,
respectively, with the methanesulfonyl radical. Table 2 collects
their corresponding CCSD(T)/6-311++G(d,p)// B3LYP/6-
311++G(d,p) ∆U, ∆H, ∆S, ∆G(gas phase), and ∆Gsol(CH2-
Cl2 solution) values, and Figure 3 shows the associated energy
profiles.
transfer), usually higher primary kinetic deuterium isotopic
effects are reported.44
As we shall show below, our ab initio calculations do confirm
that the sulfonyl radical addition to an alkene is a fast process
at room temperature. This suggested, apart from the possible
intervention of mechanism B, the operation of mechanism G
•
shown in Scheme 11. As additions of RSO2 to alkenes are
The abstraction of an allylic hydrogen from propene by CH3-
SO2 proceeds through a transition structure (see TS1abs in
relatively facile, the intermediate tertiary alkyl radicals 11 so-
obtained might abstract a hydrogen atom from the alkenes and
generate the corresponding allyl radicals 12 in exergonic
•
Figure 1) that implies a substantial barrier height (∆Gsol ) 27.8
kcal/mol). The abstracted hydrogen remains at the midway
between the methyl group in propene (C‚‚‚H ) 1.315 Å) and
(44) See e.g.: (a) Westheimer, F. H. Chem. ReV. 1961, 61, 265-273. (b) Fraser,
R. R.; Champagne, P. J. Can. J. Chem. 1980, 58, 72-78. (c) Jarczewski,
A.; Schroeder, G.; Leffek, K. T. Can. J. Chem. 1991, 69, 468-473. (d)
Grzeskowiak, I.; Galezowski, W.; Jarczewski, A. Can. J. Chem. 2001, 79,
1128-1134. (e) Jarczewski, A.; Hubbard, C. D. J. Mol. Struct. 2003, 649,
287-307.
•
the oxygen atom of the CH3SO2 radical (H‚‚‚O ) 1.267 Å).
TS1abs evolves toward a weakly bound structure, wbcpost, in
which the methanesulfinic acid formed weakly interacts with
9
7788 J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
Figure 1. B3LYP/6-311++G(d,p) structures for the species involved in the abstraction reaction between propene and the methanesulfonyl radical to give
the allyl radical and methanesulfinic acid. Distances are given in Å.
•
the allyl radical (OH‚‚‚C ) 2.471 Å). The formation of this
weakly bound complex (∆H ) -0.8 kcal/mol) is entropically
disfavored (∆G ) +7.4 kcal/mol) and consequently disappears
from the Gibbs free energy profile (see Figure 3).
profile (see Figure 3). The 2-methylpropene + CH3SO2 reaction
is an endergonic process involving a change in Gibbs free energy
quite similar to that for propene + CH3SO2 (2.6 vs 2.7 kcal/
•
mol). However, the energy barrier TS2abs is 1.1 kcal/mol lower.
This theoretical prediction supports the experimentally observed
chemoselectivity that favors the reactivity of the 2-allyl
substituted alkenes over that of the 1-susbtituted ethylenes (see
the Experimental Section).
We have also considered transition structures involving
hydrogen abstractions through the sulfur atom. However, they
are about 8 kcal/mol higher in energy than the corresponding
reactions at oxygen.
•
As can be seen in Figure 3, the propene + CH3SO2
abstraction reaction is a slightly endergonic process (∆Gsol
)
2.7 kcal/mol). The allyl radical formed can react, either within
the solvent cage or after diffusing out from it, with a second
molecule of CH3SO2H giving rise to the isomerized alkene
(similar to the reactant because of the particular model alkene
employed in our calculations, namely, propene) through an
exergonic process (∆Gsol ) -2.7 kcal/mol) involving the reverse
pathway defined by a transition structure equivalent to TS1abs
.
It is interesting to note that the structures of the two transition
structures TS1abs and TS2abs involved in the abstraction
reactions of propene and 2-methylpropene with CH3SO2 are
substantially different (see Figures 1 and 2). While the abstrac-
tion process is quite similar in both cases (the C‚‚‚H and H‚‚‚O
distances in TS2abs are 1.311 and 1.278 Å, respectively, which
agree rather well with the abovementioned values for TS1abs),
a hydrogen bonded interaction involving the second oxygen
•
In the case of the 2-methylpropene + CH3SO2 reaction, a
new structure, wbcpre, appears on the PES. It is a molecular
association formed through a weak interaction between reactants.
The formation of wbcpre is a slightly more exothermic (∆H )
-1.5 kcal/mol) process than the formation of wbcpost (∆H )
-1.0 kcal/mol), but both structures are strongly disfavored
entropically, thus disappearing from the Gibbs free energy
•
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7789

A R T I C L E S
Markovic´ et al.
Figure 2. B3LYP/6-311++G(d,p) structures for the species involved in the abstraction reaction between 2-methylpropene and the methanesulfonyl radical
to give the 2-methylallyl radical and methanesulfinic acid. Distances are given in Å.
Table 2. ∆U, ∆H, ∆G (kcal/mol), and ∆S (cal/K‚mol) Values
(Relative to Reactants), as Computed at the CCSD(T)/
atom in CH3SO2• and one of the hydrogen atoms of the second
methyl group in 2-methylpropene (the S-O‚‚‚H-CH2 distance
is 2.329 Å) is only present in TS2abs
6-311++G(d,p)//B3LYP/6-311++G(d,p) Level, for the Abstraction
of an Allylic Hydrogen in the Reactions between Propene or
2-Methylpropene and the Methanesulfonyl Radicala
.
One could be tempted to conclude that the main factor
responsible for the lower energy barrier of TS2abs should be
this S-O‚‚‚H-CH2 hydrogen bonding stabilizing interaction
which does not exist in TS1abs. However, a more detailed
analysis showed that this is not the case. Indeed, an extensive
structure
∆
U
∆
H
∆
S
∆
G
∆Gsol
R1a
TS1abs
wbc1post
P1
0.0
22.2
-0.6
4.4
0.0
19.0
-0.8
2.8
0.0
-32.3
-24.4
-1.2
0.0
26.7
4.6
0.0
27.8
7.4
3.2
2.7
•
exploration of the PES for the 2-methylpropene + CH3SO2
R2b
0.0
-2.7
19.3
21.2
-0.7
6.1
0.0
-1.5
16.1
18.0
-1.0
4.5
0.0
-18.7
-36.4
-30.6
-27.3
4.7
0.0
2.2
25.0
25.2
5.2
0.0
4.7
26.7
26.9
8.5
abstraction reaction led to the location and characterization of
a second pathway which involved the transition structure TS2′abs
in Figure 2. This transition structure does not involve any
hydrogen bonding, and therefore it is very similar to the situation
found in the propene + CH3SO2• abstraction reaction (TS1abs).
Indeed, the C‚‚‚H and H‚‚‚O distances (1.315 and 1.269 Å) in
TS2′abs are virtually identical to the corresponding ones in the
case of TS1abs (1.315 and 1.267 Å). Table 2 shows that, as a
consequence of the extra stabilization due to the hydrogen
wbc2pre
TS2abs
TS2′abs
wbc2post
P2
3.1
2.6
a The standard state 1 M at 298.15 K was assumed. ∆Gsol represents the
increment in Gibbs free energy in dichloromethane (ꢀ ) 8.93). a R1 stands
•
for the reactants (propene + CH3SO2 ), and P1, for the products (CH3SO2H
•
+ allyl radical). b R2 stands for the reactants (2-methylpropene + CH3SO2 ),
and P2, for the products (CH3SO2H + 2-methylallyl radical).
9
7790 J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
Figure 3. CCSD(T)/6-311++G(d,p)//B3LYP/6-311++G(d,p) Gibbs free energy (∆Gsolv) energy profile for the abstraction reactions between propene
(2-methylpropene) and the methanesulfonyl radical to give the allyl (2-methylallyl) radical and methanesulfinic acid.
bonding interaction, the activation enthalpy associated with
TS2abs is almost 2 kcal/mol lower than that of TS2′abs. However,
the entropy contribution disfavors TS2′abs in about the same
amount. Consequently, as shown in Table 2, the activation Gibbs
free energies for TS2abs and TS2′abs are quite similar, the former
being slightly lower by 0.2 kcal/mol.
The above analysis makes clear that the hydrogen bonding
interaction present in TS2abs cannot be responsible for rendering
the hydrogen abstraction from 2-methylpropene easier than that
from propene. As the two reactions exhibit virtually the same
endergonicities (see Table 2), the thermodynamic driving
forces45 will be similar and we must conclude that the intrinsic
barrier should be governed by the differences in ionization
energies between propene (IE ) 9.73 eV)31 and 2-methylpro-
pene (IE ) 9.22 eV).31 The expression derived by Shaik and
Schlegel for the energy barrier of reactions in which a radical
abstracts an hydrogen atom from a molecule, within the
framework of the state correlation diagrams (SCD) model,46
shows that the lower the ionization potential the lower the energy
barrier, in full agreement with the experimentally observed and
theoretically predicted chemoselectivity for the reactions studied.
A natural bond orbital (NBO) analysis47 (see Figure 4) shows
that the charge transfer from the alkene to the CH3SO2• radical
is 0.34 and 0.38 e for TS1abs and TS2abs, respectively. The
(46) (a) Shaik, S. S. J. Am. Chem. Soc. 1981, 103, 3692-3701. (b) Pross, A.;
Shaik, S. S. Acc. Chem. Res. 1983, 16, 363. (c) Shaik, S. S. In New
Theoretical Concepts for Understanding Organic Reactions; Bertra´n, J.;
Csizmadia, I. G. Eds.; NATO ASI Series, Kluwer Academic Publishers:
Dordrecht, 1988.
(45) Shaik, S. S.; Schlegel, H. B.; Wolfe, S. Theoretical Aspects of Physical
Organic Chemistry; Wiley: New York, 1992.
(47) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899-926.
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7791

A R T I C L E S
Markovic´ et al.
charge-transfer configuration
will stabilize the transition structures of the two-step mechanism
as a consequence of the electrophilic character of the methane-
sulfonyl radical. The diabatic curves of the low-lying charge-
transfer
Figure 4. Lewis representations of the TS1abs and TS2abs transition
structures as arising from a natural bond orbital (NBO) analysis.
greater efficiency of the charge-transfer mechanism in the case
of 2-methylpropene is consistent with its greater chemoselec-
tivity.
and zero (ground-state) configurations
It should be stressed at this point that the above considerations
agree rather well with one of the empirical rules governing the
free-radical substitution reactions;48 namely, electron donor
substituents (like CH3 in 2-methylpropene) will enhance transfer
•
reactions by electronegative radicals (like CH3SO2 ). Sometimes,
however, the competing addition reaction is more favorable than
hydrogen transfer as shown by Scho¨neich and co-workers in
the case of polyunsaturated fatty acids.49
will interact strongly, thus giving rise to a notable reduction in
the activation energy.46,53 The calculations presented in the next
sections fully confirmed such expectations.
From the Shaik treatment’s view, one can conclude that the
high electron affinity of the methanesulfonyl radical and the
lower ionization energy of 2-methylpropene versus propene will
favor the hydrogen-transfer reaction.
The so-called polar effect (R:H‚X a R‚+H:X-) has been
invoked to explain numerous directive effects observed in
aliphatic systems,50 but scarce theoretical work has been
developed to rationalize the experimental findings. In a recent
article,51 some theoretical analysis on reactions involving thiyl
radical-mediated cleavage of allylic C-N bonds is presented.
The authors focus on the thermodynamical analysis of the
hydrogen abstraction mechanism and mention that other alterna-
tive pathways like, for example, electron transfer, were not
considered.
B. Alternative Pathways. The initially proposed mechanism3
involved the reaction between the allyl radical and the alkene
(see Scheme 4). We carried out an extensive exploration of the
B3LYP/6-311++G(d,p) PESs for the propene and 2-methyl-
propene reactions, and the results are collected in Table 3.
Geometries of the transition structures defining the different
pathways explored, which correspond to different relative
orientations of the two reactants, can be found in Figure S1 of
the Supporting Information. One important point clearly emerges
from inspection of Table 3.
The energy barriers are much higher than those for the
reactions with the methanesulfonyl radical. Indeed, Table 3
shows that for the latter the B3LYP/6-311++G(d,p) barriers
are 5-6 kcal/mol lower. This theoretical prediction reinforces
the arguments based on the experimental facts, presented in the
previous section, in favor of mechanism A + C in Scheme 4,
in which the hydrogen atom is transferred via the sulfinic acid.
On the other hand, both mechanism E in Scheme 5 and
mechanism E + G + B also mentioned in the Experimental
Section imply addition of the methanesulfonyl radical to the
Addition, electron transfer, and other a priori suitable
mechanisms have been theoretically explored by us, and the
corresponding results are presented and discussed in the next
sections.
Particularly interesting is the connection between Shaik-type
analysis, based on Marcus treatment of hydrogen-atom transfer
reactions,45,46 and the so-called polarity-reversal catalysis in-
troduced by Roberts and co-workers.52 According to these latter
authors, a notable reduction in the activation energy for the
hydrogen-atom transfer reaction
occurs when this process is carried out through a two-step
pathway according to,
(48) Tedder, J. M. Angew. Chem., Int. Ed. Engl. 1982, 21, 401-410.
(49) Scho¨neich, C.; Dillinger, U.; von Bruchhausen, F.; Asmus, K.-D. ArchiVes
of Biochemistry and Biophysics 1992, 292, 456-467.
(50) Russell, G. A. In Free Radicals; Kochi, J. K., Ed.; Wiley: New York,
1973; Vol. 1, pp 275-331.
(51) Escoubet, S.; Gastaldi, S.; Timokhin, V. I.; Bertrand, M. P.; Siri, D. J.
Am. Chem. Soc. 2004, 126, 12343-12352.
(52) (a) Paul, V.; Roberts, B. P. J. Chem. Soc., Chem. Commun. 1987, 1322-
1324. (b) Paul, V.; Roberts, B. P. J. Chem. Soc., Perkin Trans. 2 1988,
1183-1193. (c) Roberts, B. P. Chem. Soc. ReV. 1999, 28, 25-35.
(53) (a) Epiotis, N. D.; Shaik, S. S. J. Am. Chem. Soc. 1978, 100, 1-8. (b)
Epiotis, N. D. Theory of Organic Reactions; Springer-Verlag: Berlin, 1978.
(c) Mene´ndez, M. I.; Sordo, J. A.; Sordo, T. L. J. Phys. Chem. 1992, 96,
1185-1188. (d) Lo´pez, R.; Mene´ndez, M. I.; Sua´rez, D.; Sordo, T. L.;
Sordo, J. A. Comput. Phys. Commun. 1993, 76, 235-249. (e) Sua´rez, D.;
Gonza´lez, J.; Sordo, T. L.; Sordo, J. A. J. Org. Chem. 1994, 59, 8058-
8064.
where MeSO2 is said to act as a polarity-reversal catalyst.52 The
9
7792 J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
Table 3. ∆U, ∆H, ∆G (kcal/mol), and ∆S (cal/K‚mol) Values
(Relative to Reactants), as Computed at the B3LYP/
6-311++G(d,p) Level, for the Two Transition Structures
Corresponding to the Abstraction of an Allylic Hydrogen in the
Reactions between Propene (2-Methylpropene) and the
Methanesulfonyl Radical, TS1abs (TS2abs, TS2′abs), and for the
Three Possible Transition Structures Corresponding to the
Abstraction of an Allylic Hydrogen in the Reaction between
Propene (2-Methylpropene) and the Allylic Radical (2-Methyl Allylic
Radical), TS3A-C (TS4A-C)a
out above on the allyl/2-methallyl radical + propene/2-meth-
ylpropene reactions (mechanism B in Scheme 4) allows us to
safely discard such an alternative.
However, the addition-elimination mechanism E in Scheme
5 could be, according to the data in Table 4, a much more
favorable pathway when compared with the abstraction channel
which involves much higher energy barriers [about 23 kcal/
mol at the B3LYP/6-311++G(d,p) level; see Table 3].
To further assess the viability of mechanism E, we focused
on the elimination part leading to the isomerized alkene as
shown in Scheme 5. We carried out an exhaustive exploration
on the B3LYP/6-311++G(d,p) PESs, and no saddle point for
a reaction eliminating CH3SO2H from CH3SO2CH2C•(CH3)2,
the product of addition of CH3SO2• to 2-methylpropene, could
be located. The lack of an elimination pathway has also been
reported recently for the abstraction of an allylic hydrogen in
the chlorine atom reactions with alkenes.54 For these processes,
Ragains and Fynlayson-Pitts55 speculated about the possibility
that the formation of HCl was not a direct bimolecular
abstraction reaction but a rather involved addition of the chlorine
atom followed by elimination of HCl. However, ab initio
calculations failed to locate any transition structure for the
elimination channel.54
structure
∆U
∆H
∆S
∆G
∆Gsol
TS1abs
TS3A
TS3B
TS3C
17.0
20.9
20.9
20.8
13.8
19.0
19.0
18.9
-32.3
-31.1
-31.1
-31.7
21.6
26.4
26.4
26.4
22.7
27.2
27.3
27.4
TS2abs
TS2′abs
TS4A
TS4B
TS4C
15.4
16.8
20.5
20.5
20.6
12.1
13.6
18.6
18.5
18.7
-36.4
-30.6
-33.9
-34.9
-35.2
21.1
20.8
26.8
27.1
27.3
22.8
22.5
28.0
28.4
29.0
a The standard state 1 M at 298.15 K was employed. ∆Gsol represents
the increment in Gibbs free energy in dichloromethane (ꢀ ) 8.93).
Table 4. ∆U, ∆H, ∆G (kcal/mol), and ∆S (cal/K‚mol) Values
(Relative to Reactants), as Computed at the B3LYP/
6-311++G(d,p) Level, for Some Selected Addition Pathways of
the Methanesulfonyl Radical on Propene or 2-Methylpropenea
structure
∆U
∆H
∆S
∆G
∆Gsol
Therefore, the lack of an elimination pathway for CH3SO2H
from the addition product allows us to rule out the alternative
mechanism E in Scheme 5.
propene
O-addition TS5Aadd
16.7 16.8 -34.7 25.2 25.3
3.3 4.3 -34.3 12.6 14.7
P5Aadd
TS5_6Arot 4.0 4.5 -39.2 14.3 16.1
P6Aadd
S-addition TS7Aadd
P7Aadd
3.1 4.3 -33.6 12.4 14.3
6.5 7.0 -38.0 16.4 16.6
2.8 4.2 -39.3 14.0 13.5
Finally, we also checked the viability of the two-step
mechanism F which implies the formation of an intermediate
between a radical cation of the type 13 and 14 and the sulfinate
anion (see Scheme 5). An exhaustive analysis showed that the
only existing intermediates on the PESs are those arising from
the weak interaction between the methallyl radical and sulfinic
acid (wbcpost) or, in the case of the reaction of 2-methylpropene,
a weakly bound complex formed by the interaction between
2-methylpropene O-addition TS5Badd
13.7 13.8 -35.6 22.5 23.5
2.3 3.6 -34.4 12.0 14.8
P5Badd
TS5_6Brot 2.7 3.4 -37.6 12.7 15.5
P6Badd
S-addition TS7Badd
P7Badd
2.0 3.4 -32.2 11.1 13.6
5.5 6.1 -35.9 14.9 15.1
2.6 4.2 -36.8 13.2 12.7
a The standard state 1 M at 298.15 K was assumed. ∆Gsol represents the
increment in Gibbs free energy in dichloromethane (ꢀ ) 8.93).
•
the alkene and CH3SO2 (wbcpre). However, as pointed out
previously the formation of these weakly bound complexes are
endergonic processes. Furthermore, no charge separation (elec-
tron transfer) is observed. Therefore, no moieties of the type
13 or 14 are present on the PESs, and consequently, mechanism
F can be discarded.
It has been pointed out recently56 that the PCM model could
produce relatively important errors when applied to anions and
cations. On the other hand, it is also well-known57,58 the
important role played sometimes by the explicit consideration
of solvent molecules on the reaction coordinate.
To show that our conclusions above on the unsuitability of
mechanism F are not biased by a poor treatment of solvent
effects, we performed additional calculations where the explicit
influence of solvent molecules on the stabilization of species
13 and 14 and their associated transition structures was taken
into account. Furthermore, we have chosen a highly polar
solvent (water) to enhance the potential effects of the very first
solvation shell.58 PCM-UFF and PCM-UAKS models56 were
employed to estimate solvation effects on the explicitly solvated
alkene. We explored the B3LYP/6-311++G(d,p) PESs corre-
sponding to the addition reactions of CH3SO2 to propene and
2-methylpropene, respectively. Table 4 collects the results for
the addition pathways (Figures S2 and S3 in the Supporting
Information show the geometries for the transition structures
and products present in these routes).
•
Additions of the methanesulfonyl radical (see Figures S2 and
S3 in the Supporting Information where drawings for structures
TS5-TS7, P5-P7 are presented) can be carried out either
through an oxygen atom (TS5Aadd, TS5Badd) or through the
sulfur atom (TS7Aadd, TS7Badd) giving rise to the final
products: P5Aadd, P5Badd and P7Aadd, P7Badd, respectively.
Appropriate rotations (TS5_6Arot, TS5_6Brot) lead to slightly
more stable addition products (P6Aadd, P6Badd) in the case of
the O-addition. From data in Table 4 we learn that the most
stable addition products are obtained when the reaction proceeds
through the sulfur atom (P7Aadd P7Badd). From the kinetic
viewpoint, the energy barriers are substantially lower in the case
of the addition through the sulfur atom (TS7Aadd,TS7Badd). In
any case, we can conclude that the sulfonyl radical addition to
an alkene is a fast process at room temperature (∆G‡ ) 16.6
and 15.1 kcal/mol for the reactions of propene and 2-methyl-
propene, respectively).
(54) (a) Bran˜a, P.; Sordo, J. A. J. Am. Chem. Soc. 2001, 123, 10348-10353.
(b) Bran˜a, P.; Sordo, J. A. J. Comput. Chem. 2003, 24, 2044-2062.
(55) Ragains, M. L.; Finlayson-Pitts, B. J. J. Phys. Chem. A 1997, 101, 1509-
1517.
(56) Takano, Y.; Houk, K. N. J. Chem. Theory Comput. 2005, 1, 70-78.
(57) Morokuma, K.; Muguruma, C. J. Am. Chem. Soc. 1994, 116, 10316-10317.
(58) (a) Ferna´ndez, T.; Sordo, J. A.; Monnat, F.; Deguin, B.; Vogel, P. J. Am.
Chem. Soc. 1998, 120, 13276-13277. (b) Monnat, F.; Vogel, P.; Rayo´n,
V. M.; Sordo, J. A. J. Org. Chem. 2002, 67, 1882-1889.
Although this theoretical prediction would support the
plausibility of the E + G + B mechanism, the analysis carried
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7793

A R T I C L E S
Markovic´ et al.
to a radical cation and sulfinate anion: A combination of E +
F mechanisms (see Scheme 5). Radical-induced polar substitu-
tion and elimination reactions in polar media such as water have
long been known.60 Giese and co-workers61 have pointed out
that the analysis of solvent effects in â-phosphate radical
chemistry plainly shows that a heterolytic â-C,O bond cleavage
mechanism, with formation of charged species (phosphate anion
and radical cation), operates. The low polarity of the solvent
employed in our study (CD2Cl2) strongly suggests that such a
mechanism involving formation of ionic intermediates should
be highly unlikely. Furthermore, the kinetic and thermodynamic
arguments given above when discussing the explicit consider-
ation of solvent molecules allow us to safely discard such a
mechanism.
C. Kinetic Isotope Effects. Table 1 collects the kinetic
isotope effects for the 2-methylpropene + CH3SO2• reaction at
different degrees of deuteration as computed with QUIVER28
at the B3LYP/6-311++G(d,p) level. Tunneling effects were
estimated by using the zero- and small-curvature tunneling
models.62 We employed different intervals along the reaction
coordinate and various step sizes until converged values for the
transmission coefficients were obtained. Values reported in
Table 1 were computed using the interval from s ) -1.65 to s
) +1.65 bohr on the reaction path, with a step size of 0.025
bohr. A temperature of 353.15 K, the one at which the rate
constants were obtained experimentally for the isomerization
of methylidenecyclopentane into 1-methylcyclopentene, was
assumed.
Figure 5. Structure for another abstraction pathway (TS1′abs) for the
reaction between propene and the methanesulfonyl radical.
systems (2-methylpropene ‚‚‚ H2O, CH3SO2• ‚‚‚ H2O, 2-meth-
-
ylpropene•+ ‚‚‚ H2O, CH3SO2 ‚‚‚ H2O; see Figure S4 in the
Supporting Information). From the thermodynamic viewpoint,
it is clear that formation of radical cations and sulfinate anions
is disfavored by more than 40 kcal/mol with respect to the
neutral alkene and sulfonyl radical species, according to B3LYP/
6-3111++G(d,p) calculations (both geometries as well as
energetic data are provided as Supporting Information; see
Figure S4 and Table S8). From the kinetic viewpoint, an ionic
mechanism should be accompanied by a transition structure in
which the explicit presence of solvent molecules is expected to
favor charge separation. However, the transition structures
located on the B3LYP/6-3111++G(d,p) PES (see Figure S5
in Supporting Information section), including one and two
solvent molecules, showed only a slight increase in charge
separation (less than 0.01 e) between the alkene and sulfinyl
radical, as compared with TS1abs (see Figure S5).
The theoretical values reproduce qualitatively well the
experimental observation that deuteration of the allylic positions
retards the isomerization much more than deuteration at the
terminal center sp2(C) of the alkene. The kH/k6D and kH/k4D
theoretical values for the 2-methylpropene model reaction
(4.21-4.57 and 4.08-4.43, respectively) are slightly larger than
the kH/k4D experimental measurements (3.67 and 2.97) for the
It is interesting to stress at this point that a possible pathway
avoiding the stabilization problems of the charged species
involved in the two-step process F (see Scheme 5) mentioned
above is the so-called proton-coupled electron-transfer (PCET)
mechanism.17 We were able to locate a new transition structure
on the PES (see TS1′abs in Figure 5), similar to those reported
by Olivella and co-workers17b representing a pathway where
the proton transfer is coupled to an electron transfer. Examina-
tion of the spin populations shows that the corresponding values
for the carbon and oxygen atoms involved in the hydrogen atom
transfer (0.26 and 0.29, respectively) are quite similar to those
found in TS1abs (0.30 and 0.25, respectively). Consequently,
the argument that the triplet repulsion between the unpaired
electrons at these atoms becomes negligible, thus facilitating
the PCET mechanism,17e does not apply in the present case.
Furthermore, an electron localization function (ELF) analysis59
confirmed that a greater steric hindrance in TS1′abs clearly
disfavors this transition structure (see Figure S6 in Supporting
Information) which, according to B3LYP/6-311++G(d,p) cal-
culations, resulted in higher energy than TS1abs by 3.1 kcal/
mol. As a conclusion, no PCET mechanism operates in the
diphenyldisulfone-catalyzed isomerization of alkenes.
•
5 + PhSO2 reaction, but the kH/k2D theoretical prediction (1.02-
1.03) is in excellent agreement with its corresponding experi-
mental value (1.12). This good agreement constitutes additional
theoretical support to the fact that the diphenyldisulfone-
catalyzed isomerizations of alkenes studied in this work proceed
through the mechanism A + C in Scheme 4.
Conclusions
The PESs for the model reactions between propene and
2-methylpropene with methanesulfonyl radicals were explored
at the B3LYP/6-311++G(d,p) level. The energy predictions
were further improved by carrying out CCSD(T)/6-311++G-
(d,p)// B3LYP/6-311++G(d,p) single-point calculations.
Calculations show that the isomerizations of the 2-alkylalk-
1-enes are kinetically favored over those of 1-alkylethylenes.
The difference in ionization energies between propene and
2-methylpropene seems to be the main factor responsible for
that preference, according to Shaik’s SCD model for reactions
(60) (a) Beckwith, A. L. J.; Crich, D.; Duggan, P. J.; Yao, Q. W. Chem. ReV.
1997, 97, 3273-3312. (b) Zipse, H. Acc. Chem. Res. 1999, 32, 571-578.
(61) Mu¨ller, S. N.; Batra, R.; Senn, M.; Giese, B.; Kisel, M.; Shadyro, O. J.
Am. Chem. Soc. 1997, 119, 2795-2803. See also: Horner, J. H.; Bagnol,
L.; Newcomb, M. J. Am. Chem. Soc. 2004, 126, 14979-14987.
(62) (a) Truhlar, D. G.; Isaacson, A. D.; Garet, B. C. In The Theory of Chemical
Reaction Dynamics; Baer, M., Ed.; CRC Press: Boca Raton, FL, 1985;
Vol. 4, pp 65-137. (b) Liu, Y.-P.; Lynch, G. C.; Truong, T. N.; Lu, D.;
Truhlar, D. G.; Garrett, B. C. J. Am. Chem. Soc. 1993, 115, 2408-2415.
Another possible mechanism involves sulfinyl radical addition
to the alkene, followed by heterolytic fragmentation giving rise
(59) Savin, A.; Nesper, R.; Wengert, S.; Fa¨ssler, T. Angew. Chem., Int. Ed.
Engl. 1997, 36, 1808-1832.
9
7794 J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006

Diphenyldisulfone-Catalyzed Isomerization of Alkenes
A R T I C L E S
in which a radical abstracts a hydrogen atom from a molecule.
The theoretically predicted kinetic deuterium isotope effects for
the model reactions were in good agreement with the experi-
mental measurements, thus supporting a mechanism in which
the hydrogen is transferred via the sulfinic acid, RSO2H(D),
generated by the reaction of the alkene with either diradicals
arising from polysulfone or sulfonyl radicals from diphenyld-
isulfone.
Acknowledgment. This work was supported by the Swiss
National Science Foundation, the Centro Svizzero di Calcolo
Scientifico (Manno), the SOCRATES (Oviedo/EPFL) program,
and DGI (Madrid, Spain) under Project BQU-07405-C02-02.
The authors thank Prof. J. M. Lluch (Universidad Auto´noma
de Barcelona) for his helpful comments on the calculation of
tunneling effects. J.A.S. wishes to thank Mrs. Carmen Rosa
Rodr´ıguez-Ferna´ndez, Prof. E. Mart´ınez-Rodr´ıguez (Departa-
mento de Cirug´ıa, Universidad de Oviedo), and Profs. A. Largo
and C. Barrientos (Universidad de Valladolid) for their continu-
ous support while preparing the manuscript.
An alternative mechanism based on the hydrogen exchange
between alkenes involves energy barriers 5-6 kcal/mol higher
and can be ruled out. Furthermore, the lack of an elimination
pathway for a sulfinic acid (RSO2H) from the product of the
•
alkene + RSO2 addition reaction represents theoretical evidence
Supporting Information Available: Procedures for the prepa-
ration of 11, 12, 13 for the reduction of 13 into 2-methylpropene,
for the reduction of 4-bromoheptane into heptane, for the
formation of hexa-1,5-diene and 2,5-dimethylhexa-1,5-diene by
heating sulfones 16 and 17, respectively, for the deuterium
incorporation in 1-methylcyclopentane (7-d1), for the preparation
of 8, 19, 20. Examples of kinetic measurements and description
of the spectrometers used. Cartesian coordinates and absolute
energy values for all the structures located on the ab initio
potential energy surfaces. Geometric and energetic data for
additional pathways, not included in the main body of the article,
located on the potential energy surfaces of the reactions
theoretically studied. Complete ref 27. This material is available
free of charge via the Internet at http://pubs.acs.org.
for discarding other potentially plausible mechanisms for the
polysulfone- and diphenyldisulfone-catalyzed isomerization
reactions of alkenes. On the other hand, another potentially
suitable two-step mechanism based on electron transfer to
generate radical cations, followed by proton transfer to the
sulfinate anion, can be discarded as the only intermediates
located on the potential energy surface correspond to weakly
bound complexes formed through slightly exothermic but
endergonic processes with no separation of charge between the
alkene and sulfonyl radical.
The predictions based on the quantum calculations for model
reactions with the methanesulfonyl radical, propene, and 2-me-
thylpropene are in perfect agreement with experimental studies
on the diphenyldisulfone-catalyzed isomerization of meth-
ylidenecyclopentane into 1-methylcyclopentene and deuterated
analogues.
JA0579210
9
J. AM. CHEM. SOC. VOL. 128, NO. 24, 2006 7795