Reversed electron apportionment in mesolytic cleavage: The reduction of benzyl halides by SmI2

Article abstract of DOI:10.1002/chem.201500519

The paradigm that the cleavage of the radical anion of benzyl halides occurs in such a way that the negative charge ends up on the departing halide leaving behind a benzyl radical is well rooted in chemistry. By studying the kinetics of the reaction of substituted benzylbromides and chlorides with SmI₂ in THF it was found that substrates para-substituted with electron-withdrawing groups (CN and CO₂Me), which are capable of forming hydrogen bonds with a proton donor and coordinating to samarium cation, react in a reversed electron apportionment mode. Namely, the halide departs as a radical. This conclusion is based on the found convex Hammett plots, element effects, proton donor effects, and the effect of tosylate (OTs) as a leaving group. The latter does not tend to tolerate radical character on the oxygen atom. In the presence of a proton donor, the tolyl derivatives were the sole product, whereas in its absence, the coupling dimer was obtained by a S_N2 reaction of the benzyl anion on the neutral substrate. The data also suggest that for the para-CN and CO₂Me derivatives in the presence of a proton donor, the first electron transfer is coupled with the proton transfer. Reverse breakup: In the mesolytic cleavage of the radical anions of benzyl halides that are para-substituted by CN or CO₂Me groups, the halogen departs, counterintuitively, as a radical and the benzyl system carries the negative charge (see figure).

Full text of DOI:10.1002/chem.201500519

DOI: 10.1002/chem.201500519
Full Paper
&
Kinetics
Reversed Electron Apportionment in Mesolytic Cleavage: The
Reduction of Benzyl Halides by SmI2
Offir Yitzhaki and Shmaryahu Hoz*[a]
Abstract: The paradigm that the cleavage of the radical
anion of benzyl halides occurs in such a way that the nega-
tive charge ends up on the departing halide leaving behind
a benzyl radical is well rooted in chemistry. By studying the
kinetics of the reaction of substituted benzylbromides and
chlorides with SmI2 in THF it was found that substrates para-
substituted with electron-withdrawing groups (CN and
CO2Me), which are capable of forming hydrogen bonds with
a proton donor and coordinating to samarium cation, react
in a reversed electron apportionment mode. Namely, the
halide departs as a radical. This conclusion is based on the
found convex Hammett plots, element effects, proton donor
effects, and the effect of tosylate (OTs) as a leaving group.
The latter does not tend to tolerate radical character on the
oxygen atom. In the presence of a proton donor, the tolyl
derivatives were the sole product, whereas in its absence,
the coupling dimer was obtained by a SN2 reaction of the
benzyl anion on the neutral substrate. The data also suggest
that for the para-CN and CO2Me derivatives in the presence
of a proton donor, the first electron transfer is coupled with
the proton transfer.
Introduction
When SmI2 is allowed to react with benzyl halides, an elec-
tron is transferred from the SmI2 to the substrate. The electron
transfer may merge with the departure of the halide into
a single step (dissociative electron attachment). Alternatively,
the electron transfer may produce a radical anion with a finite
lifetime (of at least several molecular vibrations) and the depar-
ture of the leaving group occurs in a second step. This point
was addressed several years ago by Jensen and Daasbjerg[13]
and Saveant et al.,[14] but not within the context of SmI2.
Jensen and Daasbjerg suggested the involvement of a carboni-
um ion configuration when electron-donating substituents are
used. The contribution of this configuration is diminished
when strong electron-accepting substituents are present such
as in the case of dicyanobenzene. Saveant et al. concluded
that in the case of a strong electron-withdrawing substituent,
such as the nitro group, the reaction is stepwise with the first
step being a reversible electron transfer and the departure of
the leaving group is rate-determining. With electron-donating
substituents the reaction is most probably a concerted one,
similar to SN2 and the addition of the electron and the expul-
sion of the halide take place simultaneously.
This paper weds two important topics, the chemistry of benzyl
halides and of SmI2. The benzylic system is one of the most im-
portant model systems. The vicinity of the a-carbon to the aro-
matic p system is, most probably, the major cause for its prolif-
erated use in the exploration of a multitude of effects and phe-
nomena. Selected examples of this are the determination of
substituent parameters,[1] substituent effects on rates of reac-
tions such as SN2,[2] and steric inhibition of resonance.[3] SmI2,
which was introduced into organic chemistry much later,[4]
became a pivotal player among the electron-transfer reducing
agents due to several advantages it has over the others. Its
popularity lies in is its versatility and ability to affect chemose-
lectivity by appropriate ligation.[5]
In most of the papers reporting the reaction of SmI2 with
benzyl halides, the benzylic system was used as a model or
a vehicle to explore the landscape of SmI2 capabilities. A
number of examples of these are the lifetime of excited SmI2,[6]
the issue of inner sphere versus outer sphere electron trans-
fer,[7] as a mechanistic probe for the reactions of SmI2–water,[8]
the effect of crown ethers on SmI2 reactions,[9] the effect of
water,[10] glyme,[11] hexamethylphosphoramide (HMPA)[12] and
so on.
In this paper we report on the mechanism of the reaction of
benzyl halides with SmI2. As will be shown later, this study led
to some interesting and unexpected results.
[a] O. Yitzhaki, Prof. S. Hoz
Department of Chemistry
Bar-Ilan University
Results and Discussion
The kinetics of the reactions of substituted benzyl bromide
with SmI2 in THF [Eq. (1)] was investigated under pseudo first
order conditions, in which the concentration of SmI2 was
2.5 mm and that of the substrate was 50 mm.
Ramat-Gan, 52900 (Israel)
E-mail: shoz@biu.ac.il
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500519.
Chem. Eur. J. 2015, 21, 9242 – 9248
9242
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Rate constant for the reaction of substituted benzyl chloride
with SmI2 in THF.
Substituent
kI [sÀ1
]
kII [mÀ1 sÀ1
]
p-Me
H
m-OMe
p-Cl
m-CF3
p-CF3
m-CN
p-CN
p-CO2Me
0.0050
0.0035
0.0053
0.007
0.007
0.011
0.015
0.25
0.1
0.07
0.105
0.14
0.13
0.215
0.3
For all the substituents used, the reaction rates were fol-
lowed spectroscopically by monitoring the disappearance of
SmI2 at 619 nm. The reactions were found to be first order in
SmI2. The kinetic order in the substrate was determined for the
p-Me, p-Cl, and p-CN derivatives, as representatives of the
whole series, and was found to be also first order (Figure 1).
The first (kI) and second (kII=kI/[substrate] order rate con-
stants for all the substrates are given in Table 1. The reaction
4.9
160
8
Table 3. The Br/Cl element effect in the reaction of substituted benzyl
halides with SmI2.
Substituent
k(ArCH2Br)
k(ArCH2Cl)
p-Me
H
m-OMe
p-Cl
m-CF3
p-CF3
m-CN
p-CN
34.0
55.7
41.4
41.4
44.6
48.4
36.0
69.4
Figure 1. Kinetic order in the p-Me, p-Cl, and p-CN benzyl bromide deriva-
tives.
first and second order rate constants are given in Table 2 and
the element effect for the various substituents is given in
Table 3.
Table 1. Rate constant for the reaction of substituted benzyl bromide
with SmI2 in THF.
The element effect shows clearly that the departure of the
leaving group occurs at the rate-determining step. Interesting-
ly, this element effect is significantly smaller than that ob-
served by other investigators for benzyl halide although in dif-
ferent systems. For example, Jansen and Daasberg[13] studied
the reduction of benzyl bromide and chloride in DMF by using
1,4-dicyanonaphthalene radical anion as a donor. The element
effect in this case was 359. An element effect of 496 was re-
ported[15] for the reactions with tri-n-butylstannane in benzene
and a value of 670 was found for the same system in another
report.[16] The low element effect observed in our case may in-
dicate a relatively early transition state. The difference in the
transition state position along the reaction coordinate may
stem from the difference in the solvents used or, more likely,
from the nearby presence of the Sm3+. Namely, the triply
charged samarium cation may electrophilically assist, through
electrostatic stabilization, the expulsion of the leaving group.
Molecular beam studies show that the electron donor may use
a front side approach to the substrate,[17] unlike the anti-ap-
proach of a nucleophile in the classical SN2 reaction. This is ex-
plained as resulting from an electrostatic stabilization of the
product. It is not unlikely that the samarium cation is located
in a way that enables it to interact both with the p system and
the leaving group simultaneously as shown in Figure 2. As will
be explained later on, this transition state structure does not
apply to the p-CN and the p-COOMe derivatives.
Substituent
kI [sÀ1
]
kII [mÀ1 sÀ1
]
p-Me
H
m-OMe
p-Cl
m-CF3
p-CF3
m-CN
p-CN
0.17
0.2
3.4
3.9
4.35
5.8
0.22
0.29
0.29
0.50
0.54
17
5.8
10.4
10.8
340
of the p-CO2CH3 derivative was too fast to follow by using
stopped flow spectroscopy and was estimated to be ꢀ200 sÀ1.
The rate constants for most of the derivatives do not vary
much ((6.35Æ3) mÀ1 sÀ1). Yet, the data in the Table show that
the p-CN derivative reacts nearly two orders of magnitude
faster than most other derivatives. Thus, the two strong elec-
tron-withdrawing substituents, p-CN and p-CO2Me deviate sig-
nificantly from the rest of the substituents. The origin of this
deviation will be discussed later. We will first focus on the de-
parture step of the leaving group.
An important diagnostic tool for the participation of a leav-
ing group departure in the rate-determining step is the ele-
ment effect. Hence, we determined the rate constants for the
reaction of substituted benzyl chlorides under the same condi-
tions. The reactions were much slower, enabling us to deter-
mine also the rate constant for the p-CO2CH3 derivative. The
Chem. Eur. J. 2015, 21, 9242 – 9248
www.chemeurj.org
9243
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The data in Tables 1 and 2, ena-
bled us to calculate the Hammett
1 values of 0.66 (r2 =0.873) and
0.63 (r2 =0.705) for ArCH2Br and
ArCH2Cl, respectively (Figure 3).
We will now discuss the mecha-
nism of the p-CN and p-COOMe
derivatives. As we have mentioned
before, these two substituents de-
viate positively by orders of magni-
tude from the shallow Hammett
Figure 2. Transition state for
the electrophilically assisted
departure of the leaving
group from benzyl chloride.
Figure 4. Hammett plot of log(k) versus sÀ for the reaction of DCN radical
anion with some substituted benzyl bromides in DMF. Data from ref. [18].
Yet, in our case, the only deviating substituent is p-CN (and
the p-CO2Me, which was not included in Wayner’s study),
whereas the substituents; m-CF3, p-CF3, and m-CN do not lie
on an ascending limb but are rather in line with the electron-
donating substituents (Figure 3). It is therefore, unlikely that
the above explanations apply to the case at hand. We suggest
that, in the present case, the change in the mechanism, results
from a change in the mesolytic mode of cleavage. In principle,
regardless of whether the reaction is a dissociative electron at-
tachment or consists of two successive steps, there are two
options for electron apportionment as shown in Equation (3).
Figure 3. Hammett plots for the reaction of benzyl bromides and chlorides
with SmI2. Data points for p-CN and p-COOMe were omitted.
plot. The meaning of a convex Hammett plot is that at the
bending zone, a mechanism change takes place.[18] Thus, in the
present case, the two substituents react by a mechanism dif-
ferent from that of the other substituents. Convex Hammett
plots for benzyl halides radical anions were previously ob-
served by Jensen and Daasbjerg[13] and Huang and Wayner[19]
by using aromatic radical anions as the electron donors.
Jensen and Daasbjerg explained the deviation (by using the
valance bond model) as a change from an SN1-like mechanism
for the electron-donating substituents to a SN2-like mechanism
for the electron-withdrawing substituents. Based on product
analysis, Huang and Wayner had shown that there is a shift
from electron transfer to the SN2 mechanism [Eq. (2)].
In one mode (3a), the departing halogen carries the nega-
tive charge leaving behind a benzyl radical. In the other (3b),
the benzylic system becomes negatively charged and the halo-
gen departs as a radical. We suggest that for most of the sub-
stituents in our case the reaction is through pathway 3a,
whereas the CN and CO2Me derivatives react through 3b.
The halogen atoms, Cl and Br, are capable of stabilizing
both a radical and a negative charge. To support the suggest-
ed mechanism, we have looked for a leaving group that toler-
ates radicals to a much lesser extent than Br and Cl and there-
fore will not react through the 3b path. We assume that an
oxygen-based leaving group will serve our purpose because
the bond energy, which reflects on the stability of the radical,
is considerably higher for the OÀC bond than for the ClÀC and
BrÀC bonds (for HOÀMe, ClÀMe and BrÀMe, the bond energies
are 92, 83.7, and 70.3 kcalmolÀ1, respectively).[20] This is also in
line with the commonplace experience that oxygen is the ulti-
mate chain carrier in fires, whereas bromine derivatives, due to
the relative stability of the bromine radical, were used as an es-
sential component in fire-extinguishing compounds. A conceiv-
able candidate for such a leaving group is p-toluenesulfonate.
We have calculated[21] the OÀMe bond energy in MeÀOSO2Ph
as well as the bond energies of the halides; MeÀCl and MeÀBr.
The data is given in Table 4.
The present work differs phenomenologically from the afore-
mentioned reports in the behavior of the electron-withdrawing
substituents. This is best demonstrated by comparison with
Wayner’s work, in which the break in the Hammett plot occurs
at H and all the electron-withdrawing substituents (m-CF3, p-
CF3, m-CN, and p-CN), fit nicely on the same ascending limb
with a 1À value of 2.6 (Figure 4).
Chem. Eur. J. 2015, 21, 9242 – 9248
www.chemeurj.org
9244
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[a]
Table 4. The XÀMe bond energy [kcalmolÀ1
]
.
Table 6. First and second rate constants for the reaction of substituted
benzyl tosylates.
B3LYP/6À
31+G*
M06/AUGÀ
M06/AUG
Àcc-pVTZ
cc-pVDZ
Substituent
kI [sÀ1
]
kII [mÀ1 sÀ1
]
PhSO3ÀMe
ClÀMe
84.2
82.8
74.8
93.8
88.0
74.1
92.0
87.8
74.0
p-Cl
m-CN
p-CN
0.02[a]
0.01[a]
0.03
0.4[a]
0.2[a]
0.6
BrÀMe
[a] Calculated as EX +EMeÀEMeX
.
[a] Rate constants derived from initial rates.
mation). To enable a comparison of the p-CN derivative rate
constant with that of the two other substituents, their reaction
rate constants were determined by using the initial rates
method.
Table 5. Ab initio computed reaction energies for Equations (4) and (5).
Eq.
B3LYP/6-31+G*
M06/AUG-cc-pVDZ
M06/AUG-cc-pVTZ
4
5
À14.9
À18.0
À17.0
À26.1
À20.3
À29.3
The data presented in Table 6 clearly show that the p-cyano-
benzyl tosylate does not deviate from the two other substitu-
ents (p-Cl and m-CN). Namely, when the “inverse” mesolytic
cleavage mode is energetically too costly, no rate deviation
and no convex Hammett plot are observed.
The data in the Table and the two isodesmic reactions
below [Eqs. (4) and (5)] for which the data is presented in
Table 5 show that sulfonate is a suitable candidate for this
purpose.
This strongly supports the assumption that the positive devi-
ation exhibited by the Cl and the Br (but not by OTs) is due to
their stability as radicals and therefore their ability to exploit
the inverse mesolytic mode. Thus, in the case of the p-CN sub-
stituent, the expulsion of these halogens as radicals is faster
than their departure as anions, whereas the aversion of the to-
sylate to become a radical compels it to leave as an anion.
It is interesting to compare the reactivity of the p-CN tosy-
late with that of the corresponding halides. In Table 7 the rate
constants for the benzylic system with the three leaving
groups Br, Cl, and OTs are given.
Therefore, we have synthesized benzyl tosylate with the p-
CN substituent, and the corresponding m-CN and p-Cl deriva-
tives as reference points. The kinetics of the reactions of these
compounds with SmI2 was followed and the reaction rates de-
termined. The m-CN and p-Cl derivatives displayed an autoca-
talytic-zero order behavior, whereas the p-CN derivative gave
a clean first order reaction (Figure 5 and the Supporting Infor-
Table 7. Rate constants for the reaction of the p-CN benzyl system with
Cl, Br, and OTs (X) as leaving groups.
X
kII [mÀ1 sÀ1
]
Cl
Br
OTs
4.9
340
0.6
The data clearly show that although tosyl group is generally
considered to be a better leaving group than Cl and Br in un-
imolecular heterolytic reactions,[22] it reacts much slower than
the chloride and the bromide derivatives, further supporting
the inverse mesolytic cleavage for the two halides.
We will turn now to the autocatalytic reactions and its
cause. We have reported in the past cases of zero order-auto-
catalytic reactions of imines with SmI2.[23] These were interpret-
ed as resulting from catalysis on the surface of micro SmI3 crys-
tals. In light of the similarity between the three OTs substrates,
and the absence of such catalysis in the chloride and bromide
derivatives, it is unlikely that the present case is also a surface
catalysis case. Moreover, addition of externally prepared SmI3
enhanced the reactions, causing the autocatalytic behavior to
vanish and the reactions became first order in the disappear-
ance of SmI2. These results suggest that a displacement of the
OTs by an iodide ion of SmI3 is causing this behavior. Indeed,
Figure 5. Kinetic trace for the reaction of p-Cl benzyl tosylate with SmI2.
Chem. Eur. J. 2015, 21, 9242 – 9248
www.chemeurj.org
9245
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 8. Products obtained in the reaction of some benzyl bromides and
tosylates with SmI3.
Substrate
Product
p-CN BnBr
p-Me BnBr
m-CN BnOTs
p-CN BnOTs
Starting material
Starting material
70% m-CN BnI
100% p-CN BnI
Scheme 1.
position. The suggested mechanism is shown in Scheme 1 for
the p-CN case.
when the ArCH2X compounds shown in Table 8 were allowed
to react with SmI3 for 2 h, only the tosylates were converted to
the corresponding iodides.
Additional support for this mechanism comes from the
effect of proton donors on the reaction rates. Proton donors,
either by hydrogen bonding or by protonation on the substitu-
ent, can provide the same effect as that of Sm3+. In the ab-
sence of a proton donor, the product, as we have shown, is
the corresponding dimer. In the presence of MeOH, at concen-
trations as low as 25 mm, the product is the corresponding tol-
uene.[26] Despite the difference in the products, MeOH did not
have any effect on the reaction rates of all the substrates (see
the Supporting Information) except for the p-CN and p-CO2Me-
substituted ones. For these two substituents, a significant rate
enhancement was observed (Table 9).
The fact that the p-CN tosylate does not show the autocata-
lytic behavior, stems, apparently, from the fact that its reaction
with SmI2 is faster than its reaction with SmI3.
Of crucial importance is the question; what causes the re-
duction by SmI2 to differ so markedly from the reduction by ar-
omatic radical anions? The explanations given to the Hammett
curvature by Daasbjerg and Wayner were the degree of inclu-
sion or its absence of a carbonium like the valence bond (VB)
structure at the transition state[13] and a switch from an SN2
mechanism to an electron transfer one.[19] Whereas, in the SmI2
reactions, the cause for the rate enhancement, as we have
shown, is the change from the normal to the inverse mesolytic
cleavage mode. To be more specific, the question is, why does
our system involve the inverse cleavage mode whereas the
other systems do not? In his work on the SRN1 mechanism,
Rossi and Bunnett suggested that in a mesolytic bond cleav-
age, the negative charge will reside with the group that is
most capable of stabilizing it.[24] Yet, the benzylic systems are
identical in all of the three scenarios and only in the present
case, the cleavage direction is reversed. The most plausible
reason for the difference seems to be rooted in the nature of
the countercation. In the present study, the countercation is
Sm3+. In the cases in which the donors were electrochemically
generated, the countercation was the carrier tetrabutylammo-
nium cation. An additional clue is the fact that unlike in the
Wayner system (in which all electron-withdrawing substituents
are on the same line), in the reaction with SmI2, only the p-CN
and p-CO2Me substituents deviate from the line, whereas the
other substituents are “well behaved”. These two substituents
differ from the others by two features. The first is that these
two substituents are capable of delocalizing the negative
charge from the benzylic position onto themselves, and the
second is that they are capable of binding to the samarium
cation.[10a,25] Although it is unlikely that the electron-withdraw-
ing power of the p-CN and p-CO2Me substituents alone is suffi-
cient to overcome the electronegativity of Cl and Br, their pair-
ing with the triply charged samarium cation, apparently results
in more efficient negative charge stabilization on the benzylic
fragment than by its localization on the nucleofuge. Hence, ac-
cording to the Bunnett paradigm,[24] the mode of cleavage will
be reversed. This cannot happen with the sterically hindered
tetrabutylammonium cation, neither can it happen with the p-
CF3 substituent nor with meta substituents in which the molec-
ular p orbital is orthogonal to the one relevant to the benzylic
Table 9. Effect of added MeOH on the rates of the reaction between
SmI2 and benzyl bromide, chloride, and tosylate.
Substrate
C (MeOH) [m]
k (rel.)
p-CN BnBr
0
0.5
1
1
2.3
3.6
1
p-CN BnCl
0
0.5
1
0
3.5
6.9
1
p-CN BnOTs
p-CO2 MeBnCl
0.5
1
0
5.3
23.3
1
0.5
1
15
28.75
The suggested mechanism is, therefore, as follows. A SmI2
molecule, to which several MeOH molecules are coordinated,
approaches the substrate and the MeOH molecule forms a hy-
drogen bond with the nitrogen atom of the cyano group or
with one of the ester oxygen atoms. Subsequently, two events
follow: One is electron transfer and the other is the tightening
of the H-bond to form a covalent bond. The possible sequen-
ces of the two events are shown in Scheme 2 for the CN deriv-
ative. The proton-coupled electron transfer (PCET),[27] which
avoids the high energy corner is the most likely route. Reach-
ing directly from the hydrogen-bonded ground state to the
para quinoidic structure avoids the formation of the high
energy radical anion and the high energy protonated nitrile,
and the negative charge ends up on the methoxide anion co-
ordinated to the Sm3+ cation.
The structure in the lower right hand side of the diagram is
depicted as a quinoidic structure in which the bromine radical
Chem. Eur. J. 2015, 21, 9242 – 9248
www.chemeurj.org
9246
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ductive. This was realized rather early when it was discovered
that cyano carbon acids behave like “normal” acids (in the
Eigen sense)[29] similar to chloroform, and unlike regular carbon
acids, in which the negative charge is delocalized away from
the carbon atom onto the activating group.[30] This behavior
was also confirmed computationally.[31] Thus, the nitrogen
atom of the cyano group, unlike that of the carbonyl in the
ester, will not carry much negative charge in its frontier orbital.
This will result in a diminished binding to the samarium cation
or to a proton, at the transition state, relative to the binding to
the p-CO2Me group, and therefore to
a
lower rate
enhancement.
Scheme 2.
Conclusion
We have shown that the classical course of electron apportion-
ment in the SmI2-induced mesolytic cleavage of substituted
benzyl chlorides and bromides is not retained when para sub-
stituents capable of delocalizing the negative charge onto
themselves are used. In these cases, because of the ability of
these substituents to bind the samarium cation or to undergo
an efficient protonation within a PCET process, placing the
negative charge on the departing halide will render the reac-
tion energetically less favored. As a result, these halides will
leave as radicals and the negative charge will be delocalized
onto the benzyl-Sm3+ system. The above conclusions are
based on the specific features of the substituents deviating
from the Hammett plots, on the ability of the leaving group to
depart as an anion or as a radical, and on the effect of MeOH.
has left. However, it could also be that the reaction is stepwise
and species at this corner is the radical, which in the next step
expels the halogen radical and provides, after prototropy, the
final para-cyanotoluene. The kinetic H/D isotope effect mea-
sured in the 0.1 to 4m concentration range of MeOH(D) was
0.9Æ0.17, in line with a proton transfer between two hetero-
atoms.
As can be seen from Table 9, the effect of MeOH is inversely
proportional to the ability of the leaving group to stabilize
a radical. According to the Hammond postulate,[28] within
a family of reactions, the more endothermic the reaction is,
the later its transition state. Thus, Br, which is capable of stabi-
lizing the product radical, will have an earlier transition state
than Cl and OTs and will therefore enjoy to a lesser extent the
enhancing effect of the methanol. The interpretation of the
effect with the OTs leaving group depends on the assigned
mechanism for the OTs. Assuming that this derivative reacts in
the inverse mesolytic mode, since the tosyl radical is less
stable than the two halogen radicals, its transition state will be
achieved much later featuring a well-developed negative
charge on the benzylic unit. This, in turn will result in a much
larger rate enhancement by the added MeOH, as is indeed
shown in the Table. In this case, the fact that the p-cyanoben-
zyl tosylate rate constant is similar to that of all the other de-
rivatives that react through the normal cleavage mode, must
be fortuitous. If, as we have suggested above, the OTs deriva-
tive reacts in the normal mode, the rate acceleration by MeOH
results from a higher ability to exploit the solvation of the leav-
ing group by MeOH, through the formation of hydrogen
bonds to two oxygen atoms of the leaving group (the third
oxygen will probably be too hindered).
Experimental Section
General methods
THF was dried over Na wire, in the presence of benzophenone,
and distilled under an argon atmosphere. The freshly distilled THF
was used for all kinetic experiments as well as for the preparative
reactions. MeOH was dried according to known procedures.[32]
Water content was determined to be lower than 20 ppm. SmI2 sol-
utions were prepared as needed from a freshly prepared 0.1m THF
solution.[33] The concentration of the SmI2 solution was spectro-
scopically determined (l=619 nm; e=635). The ArCH2OTs was pre-
pared using the following procedure for p-cyanobenzyl bromide.
p-Cyanobenzyl bromide (0.49 gr, 2.5 mm) was dissolved in acetoni-
trile (50 mL) and Ag-p-toluenesulfonate (0.71 gr, 2.5 mm) was intro-
duced into the solution. After 5 h at 508C the solution was filtered,
evaporated, and extracted with diethyl ether. p-Cl BnOTs, and m-
CN BnOTs were prepared in a similar manner. The identity of the
1
products were confirmed by H (300 MHz) and 13C (75 MHz) NMR
The final point that we would like to address is the relative
rates of the p-CN and p-CO2Me derivatives. It turns out that
the p-CO2Me substituent is significantly faster than the p-CN.
Yet, both s and sÀ suggest that p-CN is a more powerful elec-
tron-withdrawing group than the p-CO2Me (s values are 0.53
and 0.64 and sÀ values are 0.63 and 1.0 for p-CO2Me and p-
CN, respectively). The reason for this discrepancy lies in the
unique nature of the electron-withdrawing capability of the
cyano group. Although it is considered a strong electron-with-
drawing group, its electronic effect, in a significant part, is in-
spectroscopies, and HRMS analyses and their melting point were
compared with the literature values.[34]
Kinetics
The kinetics of the reactions was followed by using a stopped flow
spectrophotometer in a glovebox under nitrogen atmosphere at
room temperature. The reactions were monitored at the lmax of
the SmI2 (619 nm). Whenever a proton donor was used, it was
mixed with the substrate solution. Each set of experiments was re-
peated two to three times. Within a set, each measurement was
Chem. Eur. J. 2015, 21, 9242 – 9248
www.chemeurj.org
9247
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[8] M. Szostak, M. Spain, D. J. Procter, J. Am. Chem. Soc. 2014, 136, 8459–
8466.
[9] S. Maity, K. A. Choquette, R. A. FlowersII, E. Prasad, J. Phys. Chem. A
2012, 116, 2154–2160.
[10] a) M. Szostak, M. Spain, K. A. Choquette, R. A. Flowers, D. J. Procter, J.
Am. Chem. Soc. 2013, 135, 15702–15705; b) E. Prasad, R. A. Flowers, J.
Am. Chem. Soc. 2005, 127, 18093–18099.
[11] J. A. Teprovich, Jr., M. N. Balili, T. Pintauer, R. A. Flowers, Angew. Chem.
Int. Ed. 2007, 46, 8160–8163; Angew. Chem. 2007, 119, 8308–8311.
[12] K. A. Choquette, D. V. Sadasivam, R. A. Flowers, J. Am. Chem. Soc. 2010,
132, 17396–17398.
[13] H. Jensen, K. Daasbjerg, J. Chem. Soc. Perkin Trans. 2 2000, 1251–1257.
[14] C. P. Andrieux, A. L. Gorande, J.-M. Saveant, J. Am. Chem. Soc. 1992, 114,
6892–6904 and references therein.
[15] E. V. Blackburn, D. D. Tanner, J. Am. Chem. Soc. 1980, 102, 692–697.
[16] L. W. Menapace, H. G. Kuivila, J. Am. Chem. Soc. 1964, 86, 3047–3051.
[17] B. Jia, J. Laib, R. F. M. Lobo, P. R. Brooks, J. Am. Chem. Soc. 2002, 124,
13896–13902.
[18] a) J. E. Leffler, J. Org. Chem. 1951, 16, 1785; b) O. Exner in Advances in
Linear Free Energy Relationships (Eds.: N. B. Chapman, J. Shorter),
Plenum, New York, 1972, pp. 12.
routinely repeated three times. At the end of each series, the first
measurement was repeated to ensure reproducibility within a set.
The deviation usually observed was about 5%. First-order kinetics
was analyzed by using Kinet Asyst (v. 2.2 Hi-Tech Ltd.).
General Procedure: Product preparation under conditions simi-
lar to the kinetic measurements: A freshly prepared solution of
SmI2 (0.1m) in THF was added in the glovebox to a homogeneous
solution of the ArCH2X in dry THF. The total volume of the reaction
was 16 mL, and the final concentrations were [SmI2]=50 mm and
[ArCH2X]=50 mm. After the reaction is completed, the reaction
mixture was dissolved in diethyl ether (40 mL) and washed three
times with potassium dihydrogen phosphate buffer (15 mL). After
washing by brine (20 mL) and drying over anhydrous Na2SO4, the
solvent was filtered and evaporated under reduced pressure (iso-
lated yield>95%). The dimer (ArCH2CH2Ar) was the sole product.
With the addition of one equivalent of methanol to the reaction,
the monomer (ArCH3) was isolated as the only product (yield
>95%).
[19] Y. Huang, D. D. M. Wayner, J. Am. Chem. Soc. 1994, 116, 2157–2158.
[20] Y-R. Luo, Handbook of Bond Dissociation Energies in Organic Compounds,
CRC press, New York, 2003.
Keywords: electron transfer · kinetics · radicals · reduction ·
samarium
[21] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski„ D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[1] H. H. Jaffe, Chem. Rev. 1953, 53, 191–261.
[2] a) S. R. Hartshorn, Aliphatic Nucleophilic Substitution, Cambridge Univer-
sity Press, Cambridge, 1973, and references therein. b) G. Kohnstam, Ad-
vances in Physical Organic Chemistry, Vol. 5, Academic Press, London,
New York, 1967, pp. 121–172; c) R. A. Sneen, Acc. Chem. Res. 1973, 6,
46–53; d) R. A. Y. Jones, Physical and Mechanistic Organic Chemistry,
2nd ed., Cambridge University Press, Cambridge, 1984, pp. 155–158,
and references therein.
[3] E. C. Anslyn, D. A. Doughert, Modern Physical Organic Chemistry Univer-
sity Science Books, Sausalito, CA, USA, 2004.
[4] J. L. Namy, P. Girard, H. B. Kagan, New J. Chem. 1977, 1, 5–7. For recent
reviews on the chemistry of SmI2, see: a) M. Szostak, D. J. Procter,
Angew. Chem. Int. Ed. 2012, 51, 9238–9256; Angew. Chem. 2012, 124,
9372–9390; b) B. Sautier, D. J. Procter, Chimia 2012, 66, 399–403;
c) D. J. Procter, R. A. Flowers, T. Skrydstrup, Organic Synthesis Using Sa-
marium Diiodide: A Practical Guide, The Royal Society of Chemistry,
Cambridge, 2010; d) J. M. Concellón, H. Rodríguez-Solla, C. Concellón, V.
Amo, Chem. Soc. Rev. 2010, 39, 4103–4113; e) T. Honda, Heterocycles
2010, 81, 2719–2747; f) T. Nakata, Chem. Soc. Rev. 2010, 39, 1955–
1972; g) K. C. Nicolaou, S. P. Ellery, J. S. Chen, Angew. Chem. Int. Ed.
2009, 48, 7140–7165; Angew. Chem. 2009, 121, 7276–7301; h) K. Gopa-
laiah, H. B. Kagan, New J. Chem. 2008, 32, 607–637; i) S. Ichikawa, Chem.
Pharm. Bull. 2008, 56, 1059–1072; j) V. Faugeroux, Y. Genisson, Curr.
Org. Chem. 2008, 12, 751–773; k) D. Y. Jung, Y. H. Kim, Synlett 2005,
3019–3032; l) J. M. Concellón, H. Rodríguez-Solla, Chem. Soc. Rev. 2004,
33, 599–609; m) A. DahlØn, G. Hilmersson, Eur. J. Inorg. Chem. 2004,
3393–3403; n) D. J. Edmonds, D. Johnston, D. J. Procter, Chem. Rev.
2004, 104, 3371–3403; o) M. Szostak, M. Spain, D. J. Procter, Chem. Soc.
Rev. 2013, 42, 9155–9183; p) M. Szostak, N. J. Fazakerley, D. Parmar,
D. J. Procter, Chem. Rev. 2014, 114, 5959–6039; q) K. A. Choquette, R. A.
Flowers in Comprehensive Organic Synthesis, Vol. 1 (Eds.: G. A. Molander,
P. Knochel), 2nd ed., Elsevier, Oxford, 2014, pp. 279.
[22] H. M. R. Hoffmann, J. Chem. Soc. 1965, 6762–6765.
[23] C. N. Rao, S. Hoz, J. Am. Chem. Soc. 2011, 133, 14795–14803.
[24] R. A. Rossi, J. F. Bunnett, J. Am. Chem. Soc. 1974, 96, 112.
[25] C. N. Rao, S. Hoz, J. Org. Chem. 2012, 77, 4029–4034.
[26] This shows that the dimer formation does not result from a radical
combination but rather from an SN2 reaction of the benzyl anion on
ArCH2ÀX.
[27] R. Yella, S. Hoz, Org. Lett. 2013, 15, 5262–5265.
[28] G. S. Hammond, J. Am. Chem. Soc. 1955, 77, 334–338.
[29] M. Eigen, Angew. Chem. Int. Ed. Engl. 1964, 3, 1–19; Angew. Chem.
1963, 75, 489–508.
[30] A. C. Lin, Y. Chiang, D. B. Dahlberg, A. J. Kresge, J. Am. Chem. Soc. 1983,
105, 5380.
[31] a) R. P. Bell, The Proton in Chemistry, Chapman and Hall, London, 1973,
Chapter 10; b) S. Hoz, Z. Gross, D. Cohen, J. Org. Chem. 1985, 50, 832.
[32] D. D. Perrin, W. L. F. Armarego, Purification of Laboratory Chemicals,
3rd ed., Pergamon Press, New York, 1989.
[33] P. Girard, J. L. Namy, H. B. Kagan, J. Am. Chem. Soc. 1980, 102, 2693–
2698.
[34] S. Yoh, Y. Tsuno, M. Sawada, Y. Yukawa, J. Chem. Soc. Perkin Trans. 2
1989, 7–13.
[5] R. Yella, S. Hoz, Org. Lett. 2014, 16, 3876–3879.
[6] W. G. Skene, J. C. Scaiano, F. L. Cozens, J. Org. Chem. 1996, 61, 7918–
7921.
[7] R. J. Enemærke, K. Daasbjerg, T. Skrydstrup, Chem. Commun. 1999, 343–
344.
Received: February 8, 2015
Published online on May 12, 2015
Chem. Eur. J. 2015, 21, 9242 – 9248
www.chemeurj.org
9248
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim