Intrinsic barriers of the alternative modes of mesolytic fragmentations of C-S bonds

Full text of DOI:10.1021/ja960735x

+
+
J. Am. Chem. Soc. 1996, 118, 7235-7236
Intrinsic Barriers of the Alternative Modes of
7235
Mesolytic Fragmentations of C-S Bonds
Przemyslaw Maslak* and Jay Theroff
Department of Chemistry
The PennsylVania State UniVersity
mol ) is obtained by subtracting the difference in redox potentials
between the radical anion precursor (Er) and thiophenoxide (Eox)
from the homolytic free energy (∆Gh) of the neutral precursor
of the radical anion. The homolytic free energy of the
unsubstituted analog (∆Gh(3) ) 43 kcal/mol, where 3 is
diphenylmethyl phenyl sufide) was estimated from the known
heats of formation of the diphenylmethyl and thiophenyl radicals
using the Benson group additivity approach.8 The effect of the
substituents on ∆Gh values in 1 and 2 was evaluated with the
help of competitive thermolysis reactions.9 All substituents
showed bond-weakening effects smaller than 2 kcal/mol. The
data obtained are presented in Table 1. The reversible reduction
potentials for all the nitro compounds were measured by CV,
and the reversible oxidation potentials for the thiophenoxides
were available from the literature.10 These data are also gathered
in Table 1.
UniVersity Park, PennsylVania 16802
ReceiVed March 6, 1996
Unimolecular fragmentation reactions of radical ions to
radicals and ions (mesolytic cleavages1) have been used to
experimentally probe the relationship between the kinetics and
thermodynamics of elementary reactions.2,3 An important goal
of such investigations is the quantitative evaluation of factors
contributing to the intrinsic barriers.4 In the fragmentation of
radical ions wherein the unpaired electron resides initially in a
π orbital on one side of the scissile bond, two modes of electron
apportionment are possible.5 For example, in radical anions
the heterolytic mode of fragmentation results in the transfer of
charge across the scissile bond (eq 1), while in the homolytic
mode the charge remains localized on the same moiety (eq 2).
The radical anions of 1a and 2a have very similar rates and
activation energies of fragmentation. The cleavage of 2a•- is,
however, less endergonic than that of 1a•- by nearly 8 kcal/
mol. The fragmentation of 1b•-, which has ∆Gm similar to
that of 2•-, has an activation energy that is lower than that of
2a•- by ca. 3 kcal/mol.
(1)
(2)
We present here the first quantitative comparison of the two
cleavage modes in radical anions and show that for mesolysis
of C-S bonds the intrinsic barriers of the homolytic mode are
substantially higher (by ca. 3 kcal/mol) than those of the
heterolytic mode.
It is instructive to divide the barriers for endergonic reactions,
such as these presented here, into a thermodynamic component,
due purely to the energy difference between the initial and the
final states, and the kinetic component, termed the “overhead,”
that represents the extra energy costs associated with the
reorganization of the initial state necessary to reach the transition
state and includes all electronic (preexponential) factors.4 In
general, the overhead is the free-energy barrier to the exergonic
reverse reaction, and it is a function of ∆G. If the free energy
relationship for the reaction is known, the overhead can be
explicitly replaced with the intrinsic barrier,4 i.e., one observed
for ∆G ) 0. In the absence of a well-defined free-energy
relationship, the overhead may serve as a quantitative substitute
for the intrinsic barrier in comparisons of reactions with similar
driving forces. In these terms, the homolytic fragmentation
mode of 2•- has overheads that are significantly larger than those
found for the heterolytic mode of scission in 1•- in general,
and in 1b•- in particular.
The systems investigated included radical anions of 1 and 2.
In both systems the unpaired electron was initially highly
localized on the nitrophenyl moiety, as indicated by ESR studies.
The radical anions of 1a-c underwent unimolecular fragmenta-
tion according to the heterolytic mode (eq 1), yielding the
4-nitrodiphenylmethyl radical (not directly observed) and the
corresponding thiophenoxide. Similarly, as was shown before6
by Vianello et al., and confirmed by us, 2a-b•- fragmented in
a unimolecular reaction following the homolytic mode (eq 2),
giving 4-nitrothiophenoxide and the corresponding diphenyl-
methyl radical (not directly observed). The rates and activation
parameters for these reactions have been obtained in DMF by
cyclic voltammetry (CV) and are collected in Table 1.
The thermodynamics of the cleavage reactions has been
evaluated from a thermochemical cycle:3,7 ∆Gm ) ∆Gh - 23.06-
(Eox - Er), where the free energy of mesolysis (∆Gm, in kcal/
The fragmentation reaction involves transfer of electron
density from the π system with the unpaired electron to the
space between the carbon and sulfur atoms of the scissile bond
(σ*C-S). The good overlap between the orbitals involved is,
therefore, crucial for a “smooth” transfer of electron density
accompanying the cleavage.4 The importance of this stereo-
(1) (a) Maslak, P.; Narvaez, J. N. Angew. Chem., Int. Ed. Engl. 1990,
29, 283. (b) Mu¨ller, P. Pure Appl. Chem. 1994, 1077.
(2) (a) Save´ant, J.-M. Acc. Chem. Res. 1993, 2, 455. (b) Save´ant, J.-M.
Tetrahedron 1994, 50, 10117. (c) Save´ant, J.-M. J. Phys. Chem. 1994, 98,
3716.
(3) (a) Maslak, P. Top. Curr. Chem. 1993, 168, 1. (b) Maslak, P.;
Vallombroso, T. M., Jr.; Chapman, W. H., Jr.; Narvaez, J. N. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 73.
(4) The “intrinsic barrier” denotes here the reaction barrier (overhead)
at ∆G ) 0. Our procedure does not allow for distinction of the preexpo-
nential effects from the reorganization energy component; i.e., the same
rate may be accounted for by a large preexponential factor and large ∆Gq
or by a small preexponential factor and small ∆Gq. We assume κ ) 1 in
the Eyring treatment. The distinction is possible only if the free-energy
relationship (km ) f(∆Gm)) is known. See, for example: Adcock, W.;
Andrieux, C. P.; Clark, C. I.; Neudeck, A.; Save´ant, J.-M.; Tardy, C. J.
Am. Chem. Soc. 1995, 117, 8285.
(5) (a) Maslak, P.; Guthrie, R. D. J. Am. Chem. Soc. 1986, 108, 2628.
(b) Maslak, P.; Guthrie, R. D. J. Am. Chem. Soc. 1986, 108, 2637.
(6) Farnia, G.; Severin. M. G.; Capobainco, G.; Vianello, E. J. Chem.
Soc., Perkin Trans. 2 1978, 1.
(7) For an overview, see: (a) Wayner, D. D. M.; Parker, V. D. Acc.Chem.
Res. 1993, 26, 287.
(8) (a) Benson, S. W. Thermochemical Kinetics; John Wiley: New York,
1976. (b) Benson, S. W.; Cruickshank, F. R.; Golden, D. M.; Haugen, G.
R.; O’Neal, H. E.; Rodgers, A. S.; Shaw, R.; Walsh, R. Chem. ReV. 1969,
69, 279. (c) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.;
Levin, R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1989, 17 (Suppl.),
1. (d) McMillen, D. F.; Golden, D. M. Annu. ReV. Phys. Chem. 1982, 33,
493. (e) Stein, S. E. Structure and Properties; National Institute of Standards
and Technology, Standard Reference Data Program; NIST: Gaithersburg,
MD, 1994.
(9) Competitive thermolysis of 1 or 2 vs 3 was carried out in the mixture
of decalin, styrene, and 1,4-hexadiene. The initial rates of disappearance
of sulfides (10-15% conversion) were used to obtain the relative activation
energies for homolysis. The values of ∆∆Gq obtained at 140 and 180 °C
were used to calculate the ∆Gh for the substhituted compounds.
(10) (a) Andrieux, C. P.; Hapiot. P.; Pinson, J.; Save´ant, J.-M. J. Am.
Chem. Soc. 1993, 115, 7783. Also compare: (b) Venimadhavan, S.;
Amarnath, K.; Harvey, N. G.; Cheng, J.-P.; Arnett, E. M. J. Am. Chem.
Soc. 1992, 114, 221. (c) Bordwell, F. G.; Zhang, X.-M.; Satish, A. V.;
Cheng, J.-P. J. Am. Chem. Soc. 1994, 116, 6605.
S0002-7863(96)00735-4 CCC: $12 00 © 1996 American Chemical Society

+
+
7236 J. Am. Chem. Soc., Vol. 118, No. 30, 1996
Table 1. Thermodynamic and Kinetic Data for Fragmentation of 1•-and 2•-
Communications to the Editor
d
e
g
h
i
Er b
(V)
Eox c
(V)
∆Gh
∆Gm
kmf (T)
∆Hq
∆S‡
∆G‡
m
m
m
RIa
(kcal/mol)
(kcal/mol)
[s-1 (K)]
(kcal/mol)
(eu)
(kcal/mol)
pKaj
1a
1b
1c
2a
2b
-1.127
-1.120
-1.159
-1.114
-1.110
0.100
0.400
-0.035
0.455
42.2
41.8
41.9
42.3
41.6
13.9
6.7
16.0
6.1
0.6 (276)
0.1 (217)
0.09 (276)
0.7 (276)
0.01 (217)
17.3
14.5
16.7
16.6
13.6
3
5
-3
1
16.3
13.0
17.5
16.3
14.7
10.3
6.5k
11.2
32.2
25.3
0.455
5.5
-4
a Radical anion precursors. b Reversible reduction potential (vs SCE) of the radical anion precursors in CH3CN/0.1 M tetrabutylammonium
perchlorate (TBAP). c Reversible oxidation potential of the corresponding thiophenoxide (vs SCE) in CH3CN/0.1 M Et4NBF4 (ref 10). d Free energy
of homolysis of neutral 1 or 2 at 300 K, obtained by correcting ∆Gh(3) using the relative activation parameters obtained from competitive thermolysis
(ref 9). e Free energy of mesolysis at 300 K, evaluated from the thermodynamic cycle (see text). f Rate constants for the unimolecular fragmentation
of radical ions were measured by CV in DMF/0.1 M TBAP (1a,c, 2a, from 34 to -24 °C) or in CH3CH2CN (20% v/v)/DMF/0.1 M TBAP (1b,
2b from -46 to -74 °C). g Enthalpies of activation ((1 kcal/mol). h Entropies of activation ((4 eu). i Free energies of activation at 300 K. j pKa
values of the corresponding thiophenols (for 1, ref 20) and diphenylmethanes (for 2, ref 21) in Me2SO. k Estimated (ref 20).
electronic factor has been demonstrated experimentally in
fragmentation reactions of radical anions and cations.4,11 That
factor is, however, not important in the present case.
substituent effect is much smaller; only ca. 20% of the relative
diphenylmethyl anion stabilization energy (based on pKa values)
contributes to the lowering of the activation energy of the
fragmentation process.
As shown by ESR studies carried out on radical anions of
methyl and tert-butyl nitrophenyl sufides,12 the preferred
conformation maximizes the overlap between the π system and
the σ*C-S bond. In other words, S-R substituents behave as
electron acceptors. This trend is also followed by R ) CHPh2
in 2 which has a slightly lower reduction potential than that of
1. Calculations by semiempirical methods (PM3)13 for 2•- and
ab initio methods for small model compounds14 indicate that
the lowest energy conformations in radical anions have a near
perfect dihedral angle (close to 90°) between the plane of the π
system and the scissile bond. Additionally, as shown by
semiempirical calculations, in 2•- the rotation around the CPh-S
bond has only a small barrier (ca. 2 kcal/mol). All these
considerations exclude the interference of a nonreactive con-
formation as a reason for the increased overhead in 2•- as
compared to 1•-.
This effect of the extent of charge delocalization in the
transition state is a general property of the fragmentation
reaction; i.e., the heterolytic mode of mesolysis is expected to
have a smaller overhead than the homolytic mode. The
difference in overheads should depend on the difference in the
electronegativity of the atoms of the scissile bond, i.e., the
relative ability of the fragments to accommodate the charge.
As have been qualitatively demonstrated previously,5,15,16 such
differences are large for C-O or C-N and small for C-C bond
fragmentations in radical anions. In the fragmentation of C-S
bonds presented here, both modes of cleavage are directly
observable, and the intrinsic barrier differences (ca. 3 kcal/
mol) are sufficiently large to be outside of the experimental
error.17
In light of these experiments, it is perhaps not surprising that
the large majority of radical ion fragmentations reported in the
literature2ab,18 are examples of the heterolytic mode of cleavage
and that the radical-anion coupling reactions (that are a
component of many SRN1 reactions2ab,19) are formally hetero-
genic processes (reverse of eq 1). These intrinsic barrier
differences should be explicitly included in any free-energy
relationship for mesolytic fragmentation.
The difference in reactivity between the two modes of
fragmentation may be traced to the delocalization of charge in
the transition state. The shift of the electron density into the
σ*C-S bond is easily accommodated in 1•-, with the thiophenyl
moiety being able to stabilize the charge quite well. In 2•-
,
the diphenylmethyl moiety is a much poorer electron-density
acceptor. As the result, the transition state for cleavage of 1•-
has the “excess” negative charge delocalized over the entire
molecule, while the delocalization of charge in 2•- is largely
limited to the nitrophenylthio moiety.
Acknowledgment. This research has been sponsored by a grant
from the NSF.
These delocalization effects are clearly visible in the way
the substituents affect the activation energy of fragmentation.
In 1•- the change in the activation energy upon substitution
follows the same trend as one observed for the pKa values of
the corresponding thiophenols. Indeed, ca. 70% of the relative
thiophenoxide stabilization energy due to the substitution
(2.3RT∆pKa) is reflected in the transition state. In 2•- the
JA960735X
(15) (a) Guthrie, R. D.; Patwardhan, M.; Chateauneuf, J. E. J. Phys. Org.
Chem. 1994, 7, 147. (b) Wu, F.; Guarr. T. F.; Guthrie, R. D. J. Phys. Org.
Chem. 1992, 5, 7.
(16) (a) Maslak, P.; Narvaez, J. N.; Kula, J.; Malinski, D. J. Org. Chem.
1990, 55, 4550. (b) Maslak, P.; Kula, J.; Chateauneuf, J. E. J. Am. Chem.
Soc. 1991, 113, 2304.
(17) Although the absolute values of the thermodynamic estimates may
be in error by 2-4 kcal/mol, the errors of the relative data are estimated to
be less than 1 kcal/mol. The comparison of differences between the modes
of cleavage is based on the relative quantities.
(18) For a review, see: (a) Chanon, M.; Rajzmann, M.; Chanon, F.
Tetrahedron 1990, 46, 6193. (b) Save´ant, J.-M. New J. Chem. 1992, 16,
131.
(11) See, for example: (a) Gan, H.; Leinhos, U.; Gould, I. R.; Whitten,
D. G. J. Phys. Chem. 1995, 99, 3566. (b) Perrott, A. L.; Arnold, D. R.
Can. J. Chem. 1992, 70, 272. (c) Norris, R. K.; Barker, S. D.; Neta, P. J.
Am. Chem. Soc. 1984, 106, 3140. (d) Meot-Ner (Mautner), M.; Neta, P.;
Norris, R. K.; Wilson, K. J. Phys. Chem. 1986, 90, 168. (e) Maslak, P.;
Narvaez, J. N.; Vallombroso, T. M., Jr. J. Am. Chem. Soc. 1995, 117, 12373.
(f) Maslak, P.; Chapman, W. H., Jr.; Vallombroso, T. M., Jr.; Watson, B.
A. J. Am. Chem. Soc. 1995, 117, 12380.
(19) (a) Bunnett, J. F. Acc. Chem. Res. 1978, 11, 413. (b) Kornblum, N.
Angew. Chem., Int. Ed. Engl. 1975, 14, 734. (c) Rossi, R. A.; de Rossi, R.
H. Aromatic Substitution by the SRN1 Mechanism; ACS Monograph 178;
American Chemical Society: Washington, DC, 1987. (d) Julliard, M.;
Chanon, M. Chem. ReV. 1983, 38, 425. (e) Chanon, M.; Tobe, M. L. Angew.
Chem., Int. Ed. Engl. 1982, 21, 1.
(12) Alberti, A.; Guerra, M.; Martelli, G.; Bernardi, F. Mangini, A.;
Pedulli, G. F. J. Am. Chem. Soc. 1979, 101, 4627.
(13) The semiempirical MNDO-PM3 calculations (Stewart, J. J. P. J.
Comput-Aided Mol. Des. 1990, 4, 1) were carried out using Spartan 4.0
(Wavefunction Inc.).
(20) Bordwell, F. G.; Hughes, D. L. J. Org. Chem. 1982, 47, 3224.
(21) Zhang, X.; Bordwell, F. G. J. Org. Chem. 1992, 57, 4163.
(14) Maslak, P.; Hoag, A. Unpublished results.