Nucleophilic displacement versus electron transfer in the reactions of alkyl chlorosilanes with electrogenerated aromatic anion radicals

Article abstract of DOI:10.1016/j.electacta.2015.01.182

Anion radicals of a series of aromatic compounds (C₆H₅CN, C₆H₅COOEt, anthracene, 9,10-dimethyl-, 9,10-diphenyl-and 9-phenylanthracene, pyrene and naphthalene) react with trialkyl chlorosilanes R¹R²R³SiCl (R^1-3 = Me, Et; R^1,2 = Me, R³ = t-Bu) in multiple ways, following classical bimolecular schemes. The ratio of one-electron transfer (ET) to a two-electron process (S_N2-like nucleophilic attack of the reduced form of mediator on the chlorosilane, with k₂ ? 10²-10⁸ M^-1 s^-1) is inversely related to the steric availability of Si for nucleophilic displacement reactions. The nucleophilic substitution pathway mainly results in mono-and disilylated aromatic products. Paralleling the electrochemical data with DFT calculations, the role of silicophilic solvent (DMF) in S_N process was shown to be quite complex because of its involvement into coordination extension at silicon, dynamically modifying energetics of the process along the reaction coordinate. Although 2,2'-bipyridine also forms delocalized persistent anion radicals, they do not induce neither ET nor S_N reactions in the same manner as aromatic mediators. Silicophilicity of 2,2'-bipyridine being superior to that of DMF, a R₃SiCl·bipy complex of hypercoordinated silicon with electroactive ligand was formed instead, whose reduction requires about 1 V less negative potentials than bipyridine itself.

Full text of DOI:10.1016/j.electacta.2015.01.182

Electrochimica Acta 158 (2015) 457–469
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Nucleophilic displacement versus electron transfer in the
reactions of alkyl chlorosilanes with electrogenerated aromatic
anion radicals
Saida Soualmi 1, Mamadou Dieng 2, Ali Ourari 3, Diariatou Gningue-Sall 2,
,4
*
Viatcheslav Jouikov
Molecular Chemistry and Photonics, UMR 6226, University of Rennes 1, 35042 Rennes, France
A R T I C L E I N F O
A B S T R A C T
Article history:
Anion radicals of a series of aromatic compounds (C6H5CN, C6H5COOEt, anthracene, 9,10-dimethyl-,
9,10-diphenyl- and 9-phenylanthracene, pyrene and naphthalene) react with trialkyl chlorosilanes
R1R2R3SiCl (R1-3 = Me, Et; R1,2 = Me, R3 = t-Bu) in multiple ways, following classical bimolecular schemes.
The ratio of one-electron transfer (ET) to a two-electron process (SN2-like nucleophilic attack of the
reduced form of mediator on the chlorosilane, with k2 ffi102-108 MÀ1 sÀ1) is inversely related to the steric
availability of Si for nucleophilic displacement reactions. The nucleophilic substitution pathway mainly
results in mono- and disilylated aromatic products. Paralleling the electrochemical data with DFT
calculations, the role of silicophilic solvent (DMF) in SN process was shown to be quite complex because of
its involvement into coordination extension at silicon, dynamically modifying energetics of the process
along the reaction coordinate. Although 2,2'-bipyridine also forms delocalized persistent anion radicals,
they do not induce neither ET nor SN reactions in the same manner as aromatic mediators. Silicophilicity
Received 12 November 2014
Received in revised form 25 January 2015
Accepted 28 January 2015
Available online 30 January 2015
Keywords:
Chlorosilanes
aromatic anion radicals
silicophiles
electrochemical reduction
redox catalysis
of 2,2'-bipyridine being superior to that of DMF, a R3SiCl bipy complex of hypercoordinated silicon with
Á
electroactive ligand was formed instead, whose reduction requires about 1 V less negative potentials than
bipyridine itself.
ã 2015 Published by Elsevier Ltd.
1. Introductione
possibility of both single electron transfer (ET) and a pure
nucleophilic pathway in nucleophilic reactions of chlorosilanes
Silyl protection-deprotection methodology as well as many
synthetic procedures of organosilicon compounds are based on
nucleophilic reactions exploiting large versatility of electrophilic
properties and steric bulk of R3SiY (Y = Hal, AlkO, MeCOO, R2N, CN
etc) precursors [1,2]. With all importance of the chemistry of
introduction and removal of silyl groups, it is surprising that
relatively few quantitative studies on the fundamentals of
reactivity of chlorosilanes are available [2]. Corriu evoked the
[3]. The first one involves one-electron transfer from an electron-
rich nucleophile to an electrophilic chlorosilane, supposedly with
the intermediacy of an anion radical of the chlorosilane [4]. Since
the reduction of chlorosilanes requires very negative potentials [5]
and—at least of those with aliphatic substituents—occurs via
dissociative ET mechanism [6,7], the ET pathway seems less
probable, though in the case of aryl chlorosilanes this possibility
cannot be ruled out [7,8]. On the other hand, it is well documented
that heterogeneous ET to alkyl chlorosilanes induces cleavage of
the Si—Cl bonds leading to the reduction products with Si—Si or
Si—H bonds [9].
Direct cathodic reduction of chlorosilanes, pioneered by
Bobersky [10], Hengge [11] and Corriu [5,12], includes two-
electron cleavage of the Si—Cl bond resulting in silyl anions. The
nucleophilic attack of these latter on the starting molecule or on
other chlorosilane, added to the solution [13–15], leads to stable
final products via the formation of a Si—Si bond (Eq. (1)). The
* Corresponding author. Tel.: +33 22 323 6293; fax: +33 22 323 6955.
E-mail address: vjouikov@univ-rennes1.fr (V. Jouikov).
Permanent address: Faculté des Sciences et de la Technologie et Sciences de la
1
Matière, Université de Tiaret, 14000 Tiaret, Algeria.
2
Permanent address: LCPOAI, Chemistry department, University Cheikh Anta
DIOP of Dakar, BP5005 Dakar, Senegal.
3
Permanent address: Laboratory of Process Engineering, University of Setif,
19000 Setif, Algeria.
4
ISE member.
http://dx.doi.org/10.1016/j.electacta.2015.01.182
0013-4686/ã 2015 Published by Elsevier Ltd.

458
S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
intermediacy of silyl radicals and radical pathway of formation of
disilanes were ruled out already in earlier works: the addition of
PhOH during the electroreduction of Ph3SiCl lead to the quantita-
tive protonation of triorganosilyl anion and no disilane products
were formed under these conditions [5], Eq. (2).
by virtue of coordination extension ability of Si, the formation of
penta- or hexa-coordinate intermediates. Depending on whether
Berrypseudorotationoccursornotinsuchtransients,itresults inthe
products with retention or inversion of configuration [1]. Moreover,
nucleophilic reactions at silicon are often promoted by silicophilic
co-reagents such as N,N-dimethylaminopyridine, HMPA or DMF
[1,38] that induce pentacoordination at Si prior its interaction with
the proper nucleophile (Eq. (3)). Silaphilic additives such as DMPU
(tetrahydro-1,3-dimethyl-2(1H)-pyrimidinone),HMPAorDMFwere
commonly used in organosilicon electrosynthesis and voltammetry
[32,33,36,39–42] though their effect on the process studied is quite
difficult to quantify.
(1)
Ph OH
2Ph3Si Cl
!
Ph3SiH þ Ph3SiO Ph
(2)
þ2e
(3)
Given the wide use of chlorosilanes as protecting groups in
organic and organosilicon chemistry, examination of the reactions
of electrogenerated anion radicals with alkyl chlorosilanes might
provide an important insight into the ratio between electron
transfer and nucleophilic substitution (SN) pathways in their
reactivity. In the present paper, we checked the feasibility and
efficiency of redox mediated reduction of model alkyl chlorosilanes
versus silylation of aromatic mediators in view of assessing
the synthetic potential of these processes for preparing the
corresponding organosilicon products.
The reduction of chlorosilanes looks therefore quite similar
to that of haloalkanes [16]. Dissociative electron transfer
triggering the reduction of alkyl chlorosilanes requires about
1.5 V more negative potentials than the reduction of silyl
radicals to the corresponding silyl anions, which totally bans
the occurrence of any radical reactions in this process. Indeed,
the reduction of the majority of chlorosilanes is usually
observed at À3.2 . . . À2.4 V vs. SCE [9], whereas the E1/2 of
the couple Ph3Siꢀ/Ph3SiÀ, estimated by photomodulated voltam-
metry, is À1.39 V vs SCE [17]; this latter also agrees with the
generation of silyl radicals from the oxidation of silyl anions at
close potentials, as shown by real-time EPR-spectroelectrochem-
istry [18]. Thus this fundamental limitation precludes not only
the intermediacy of silyl radicals and radical-based reactions in
the direct cathodic reduction of R3SiCl compounds but also the
use of many electrophilic reagents susceptible to react with
R3SiÀ anions because these electrophiles undergo own reduction
at less negative potentials. A possible solution for debugging this
situation could be the use of redox mediators, e.g. aromatic
anion radicals, which would convey homogeneous character to
the process and expectedly reduce the applied potentials by up
to 0.6 V. Since first reports in mid-seventies on the reduction of
aliphatic halides by anion radicals of aromatic hydrocarbons—
homogeneous redox catalysis [19–21],—a great progress was
accomplished in this field as summarized in the comprehensive
review [22]. An analytical approach for such reactions has been
developed by Savéant et al. using direct consideration of
homogeneous redox catalysis [23–26] and by Lund and
Daasbjerg using competition method between ET and coupling
of alkyl radicals with the mediator [27–30].
2. Experimental
A PAR-2273 and an EG&G 362 potentiostats were used for
voltammetry and large-scale electrolyses, respectively. For cyclic
voltammetry, a 2 mm in diameter glassy carbon (GC) and a 1 mm Pt
disk working electrodes were used. A 2.5 Â 50 mm GC rod,
separated from the analyte by a sintered glass diaphragm, was
used as counter electrode. Peak potentials Ep were measured
relative to Pt wire electrode electrochemically covered with
polypyrrole and corrected using ferrocenium/ferrocene reversible
system (E0Fc+/Fc(DMSO) = 0.31 V vs. SCE [43]). The working
electrode was carefully polished with Struers FEPA P 4000 paper
and rinsed consecutively with acetonitrile, ethanol and diethyl
ether before each run. Large-scale electrolyses were carried out in a
40 mL two-compartment cell fitted with a 30 Â 20 Â 0.5 mm Pt
plate cathode and a 2.5 Â75 mm GC rod anode.
ChromatographiccontroloftheelectrolysesandGC–MSanalyses,
providing the distribution of products before any separation or a
macro work-up, were performed using an HP-5973 MSD (EI mode,
70 kV) apparatus (Agilent Technologies) with a 0.25 mm  30 m
capillary column (OPTIMA-240 from Mackerel-Nägel).
Meanwhile,
allyl, vinyl [32] and alkyl [33]), electrogenerated via the reductive
cleavage of C—X (X Cl, Br, I) bonds at less negative potentials than
s-carbanions of different nature (benzyl [31], aryl,
DMF was used as a solvent (kinetic measurements) or was added
toTHF(approx.1/10v/v)forlarge-scaleelectrolyses.Analytical grade
DMF (Aldrich) was passed through a columnwith vacuum-activated
(at 150 ꢁC) neutral Al2O3. CH3CN and THF were distilled before using
from CaH2 and from sodium benzophenone ketyl, respectively.
Et3SiCl and M3SiCl were kept over Mg turnings and distilled prior to
the electrolyses. The supporting electrolytes, Bu4NPF6 or Bu4NBF4
(Aldrich), activated in vacuum at 80 ꢁC for 10 hours and kept in a
vacuum desiccator over P2O5, were used as 0.1 M solutions. The
mediators used, all from ACROS, were used as received.
2,2'-Bipyridine (Aldrich) was additionally sublimed before using.
All experiments were carried out under an inert Ar atmosphere.
Typically, preceding the main process, a pre-electrolysis was
carried out for in situ drying the solution, as was suggested by Biran
[44]. For this, the solution of 1 mmol of Me3SiCl in 20 mL THF/0.1 M
¼
those of the reduction of chlorosilanes, act as efficient
C-nucleophiles towards the chlorosilanes themselves. These
reactions were reported to result in silylated products with
Si—Csp3 bond. Therefore in order to minimize such SN2-like
nucleophilic interactions, similar to those between aromatic anion
radicals and alkyl halides [34], and to act as outer-sphere
mediators, the anion radicals must be rather of
p-type, quite
delocalized and weakly nucleophilic. So far, only few accounts
were available on homogeneous reduction of germanium [35] and
silicon chlorides [36,37] using such
p-anion radicals.
Ontheotherhand,inspiteofsomesimilaritiesoutlinedabove,the
patterns of bimolecular substitution mechanisms at C and Si are
quite different: while SN2 in carbon chemistry is a classical example
of this type of processes, nucleophilic displacement at Si involves,

S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
459
Bu4NBF4, containing the mediator and 1–2 mL of DMF, was
electrolyzed at –1.0 V for 1.5 h to reduce the protons arising
from the hydrolyses of the chlorosilane by residual water. After the
pre-electrolysis was over, the target chlorosilane (10 mmol) was
added through the syringe and the potential of the working
electrode was switched to the Ep of the mediator used. At the end of
the electrolysis, the catholyte was quenched with saturated
aqueous solution of NaCl in order to transform unreacted
chlorosilane into the corresponding R3SiOSiR3. The mixture was
extracted by cyclohexane, the organic phase was then dried over
Na2SO4, concentrated and filtered through a 3 cm SiO2 cartridge.
The products were analyzed by GC–MS by comparing with the
authentic samples.
ZPE-corrected energies, NBO charges and profiles of potential
energy have been calculated using Gaussian 03 suite [45] on DFT
B3LYP/Lanl2DZ non-zero frequency optimized structures.
Fig. 1. (a) Reduction of Et3SiCl (5 mmol LÀ1
)
and (b) the voltammograms of
3. Results and discussion
electroreduction of naphthalene (10À2 mol LÀ1) at a GC disk electrode upon addition
of Et3SiCl (down the arrow: 0, 2, 4, 8 and 16 mmol LÀ1). Solvent DMF + 0.1 M
3.1. General considerations
Bu4NPF6. Scan rate
v
= 500 mV sÀ1, T = 20 ꢁC.
Several organic mediators (P) that form, upon one-electron
uptake, relatively stable delocalized anion radicals (Q) and ranked
according the reducing power that they provide to homogeneous
electron transfer to chlorosilanes were considered (Table 1).
9,10-Dimethyl anthracene and benzonitrile were previously used
for proving dissociative character of the cleavage of Si—Cl bond in
Me3SiCl via its indirect electroreduction. The synthetic outcome of
this process was not studied then [36]. Three monochlorosilanes
R1R2R3SiCl (R1À3 = Me; R1À3 = Et, and R1,2 = Me, R3 = t-Bu), widely
used both in silyl protection and in synthesis of silicon organic
products, were considered as substrates all showing single
irreversible two-electron reduction steps at À3.10, À2.87 and
À2.82 V vs. SCE, respectively.
All mediators listed in Table 1 show the catalytic effect when a
chlorosilane is added to the solution. As exemplified in Fig.1 in case
of naphthalene, the one-electron reduction current (ipd) of the
mediator increases upon progressive addition of Et3SiCl; at
the same time the oxidation current of naphthalene anion
radicals decreases up to its total disappearance (at the ratio
[Et3SiCl]/[P] ꢂ 12/1). It is seen that at this point the normalized
d
Fig. 2. Normalized kinetic currents ipcat/ip of the reduction of: ( ) naphthalene,
4
(*) PhCN, (&) PhCOOMe and (ꢃ) anthracene (CP = 10À3 mol LÀ1 in DMF/0.1 M
Bu4NPF6) in the presence of Et3SiCl as a function of Et3SiCl-to-mediator ratio.
v = 500 mV sÀ1, T = 20 ꢁC.
d
kinetic current ipcat/ip of the mediator only reaches the value
of 2; other mediators show quite similar behavior (Fig. 2). This
indicates that the reaction between the reduced form Q and the
chlorosilane is not limited to pure electron transfer and includes
a substrate-mediator addition [46,47]; the contribution of this
side process limits the propagation of the catalytic scheme
(Eq. (5b)).
P + e @ Q
(4)
Table 1
Aromatic redox mediators and the rate constants (MÀ1 sÀ1) of the reactions of their anion radicals with chlorosilanes: electron transfer (ET)a vs. nucleophilic displacement
(SN)
Mediator
ÀE0P/Q, Vb
Me3SiCl
k1 (ET)
Et3SiCl
k1 (ET)
t-BuMe2SiCl
k1 (ET)
k2 (SN)
k2 (SN)
k2 (SN)
Triphenylphosphine
Biphenyl
Naphthalene
2.65
2.55
2.50
2.23
2.18
2.10
2.02
1.90
1.89
1.83
1.79
4.78 Â 107
1.69 Â108
5.61 Â106
1.34 Â106
6.32 Â105
6.87 Â103
2.74 Â103
2.39 Â108
7.01 Â107
3.59 Â107
5.21 Â105
2.12 Â105
6.92 Â106
1.18 Â 105
5.13 Â104
3.02 Â107
6.60 Â105
2.94 Â105
1.04 Â105
1.38 Â 103
5.79 Â102
3.25 Â106
5.49 Â104
2.37 Â104
Benzonitrile
Methylbenzoate
2,2'-Bipyridinec
Pyrene
9,10-Dimethylanthracene
Anthracene
2.96 Â103
2.92 Â102
2.39 Â102
70.74
1.79 Â104
1.78 Â 103
1.45 Â103
4.25 Â102
1.21 Â102
9.37 Â103
31.30
3.05
2.50
0.74
0.32
1.35 Â103
1.32 Â102
1.08 Â 102
31.57
7.89
2.11
5.67 Â102
1.43 Â102
9-Phenylanthracene
9,10-Diphenylanthracene
13.65
a
b
c
¼
With
V, vs SCE; DMF/0.1 M Bu4NBF4.
Falls out of the reaction series, see text.
DG0 obtained using the DGSi-Cl from B3LYP/Lanl2DZ frequency analysis.

460
S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
between kf and the processes with k1, k2, and k6 should be
considered.
þe
P @ Q
(4)
Under such circumstances, it is difficult to check whether
ET (1e):
primary R3Siꢀ radicals undergo second ET (either homo- or
heterogeneous), or dimerization or else other radical reactions
with the components of the solution. This situation can be
simplified considering the following. With the mediators used in
this study, the primary product R3SiQ (Eqs. (5b), (10) and (11)) may
only contain Si—C or Si—O bonds, both stable enough for not
allowing any reverse reactions (supposing R3SiQ to be kinetic
product) that might lead to the disilane (thermodynamic)
products. Though radical-radical dimerization is very fast and
usually occurs at the diffusion limit (for Me3Siꢀ, k = 3 Â109 MÀ1 sÀ1
[48] and for Et3Siꢀ, k = 1010 MÀ1 sÀ1 [49]), the efficient reducing
potential in the reaction space is such that the silyl radicals5 (with
kf
R3SiCl þ DMF @ R3SiðDMFÞCl
(7)
k1
_
Q þ R3SiClÀ!ClÀR3Si þ P
(5a)
kdim
_
2R3Si ! R3SiSiR3
(8)
(9)
k3
R3Si þ Q ! R3SiÀ þ P
_
own reduction potentials
E
ꢄ À0.5 . . . À1.4 V [17,18]) are
instantaneously transformed into silyl anions, which is consistent
with the results of Corriu [12]. It is quite obvious then that if a
disilane is formed in the process starting with eqs. ((1), (4), (5a),
it could only have arisen from the reactions following the
reduction of R3Siꢀ radicals (eqs. (6), (6a)). An SN-like nucleophilic
pathway is therefore assumed the only one accounting for the
formation of Si-Si products.
k4
À
_
R3Si þ Q ! R3SiQ
(10)
SN (2e):
k2
Q þ R3SiCl ! R3SiQÁ
(11)
(12)
ÀClÀ
þe
À
_
R3Si ! R3Si
(6)
Q
R3SiQÁ! R3SiQÀ þ P
R3SiCl
ÀClÀ
k5
R3SiÀ
!
R3SiSiR3
(6a)
2R3SiQÁk!dim2 R3SiQQSiR3
(13)
(14)
In an in situ electrochemically dried solution (when the
protonation of silyl anions is minimized), the nucleophilic attack
of R3SiÀ (Eq. (6a)) on starting R3SiCl is faster compared to its
R3Si Cl
R3SiQÀ
!
R3SiQSiR3
k6
addition across the C O or CRN bonds [1,4] of the most
¼
electrophilic among the used mediators. Therefore, with R3SiÀ
produced in the solution through Eqs. (4)–(6), the yield of the
corresponding disilane can roughly be considered as reflecting the
contribution from pure ET, Reaction (5a), k1.
Two competing electron transfers (ET) trigger this sequence: a
one-electron transfer, Eq. (5a) with ensuing radical coupling
(Eq. (8), kdim), and a two-electron (SN2-like) attack of the mediator
anion radical (by its site with highest negative charge localization)
on the chlorosilane (Eq. (11)). The steps with k1 and k2 are
supposed to be chemically irreversible since both involve Si—Cl
bond cleavage (plus Q—Si bond formation, Eqs. (10), (11).
Admitting steady-state for [R3Siꢀ] in the ET scheme implies that
k1 is rate-limiting, i.e. k1 << (kdim + k3 + k4), and the concentration
of silyl radicals is manly determined by the competition between
k1 and their consumption in Reactions (8)–(10). The competition
with the SN-process (Eqs. (11)–(14) brings k2 into the kinetic
equation. Thus, without taking into account diffusion of reagents
and the contribution of direct reduction of radical species at the
cathode due to homogeneous character of the process, the rate law
is
The anion radical of the mediator does not solely act as an
outer-sphere electron donor towards Si-Cl substrates and the
process involves several successive steps. Considering the process
in its whole complexity, - involving multiple instable species and
their interconversion depending on the reaction conditions, - in
the absence of data on the coordination state of silyl intermediates
and their interactions with various nucleophiles in the solution
seems a non-realistic task, at least at the present. In fact, it would
involve too many adjustable variables for obtaining reliable results.
Therefore, we reduced the case to a simplified model, neglecting
several seemingly secondary reactions (e.g. protonations and
radical coupling of secondary radicals) and based on steady-state
assumption for major intermediates (which holds sound in most
practical situations [26]), developed for homogeneous reduction of
alkyl halides [23]. In the following principal steps, provided an
over-equimolar presence of DMF prior any ET, we assume
[R3SiCl] = [R3Si(DMF)Cl] (Eq. (3)). Otherwise, the competition
Àd½QŠ=dt ¼ ½QŠððk1 þ k2Þ½R3Si ClŠ þ ðk3 þ k4Þ½R3SiŠ
þ k5ð½R3SiQŠÞ
(15)
Steady-state for [R3Siꢀ] (Eq. (16) implies that kdim[R3Siꢀ]2 ! 0, even
if kdim is close to the diffusion limit
2
d½R3SŠ=dt ¼ k1½QŠ½R3Si ClŠ À kdim½R3SiŠ À ðk3 þ k4Þ½QŠ½Š½R3SiŠ ¼ 0
(16)
5
Á
Dimerization of homologous Ph3Ge radicals was postulated in [35]. However
provided E
¼ À0:61V vs SCE [17], an increase in the driving
Þ
ð1=2ðPh3 GeÁ=Ph3GeÀ
force of their reduction to germyl anions at the potential of E = À2.1 V,
required for the reductive cleavage of Ge—Cl bond, is by 5 Â 1012 times at
the level of ks. Most probably, no radical coupling could therefore compete
with their reduction under these conditions.
½R3SiꢀŠ ¼ k1½R3Si ClŠ=ðk3 þ k4Þ
(17)

S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
461
the zone of diffusion control by the back process was attained [25],
0
À
so simple determination of EðR SiCl=R SiꢀþCl by extrapolation of the
Þ
3
3
line with 1/58 mV slope to log(kD) is not possible. However, some
estimation concerning their values can be made here. First,
Cabasso et al. [8] reported that Ph2SiCl2 and MePhSiCl2 show, at
T = -41 ꢁC, reversible reduction with E0app (as an average of Epc and
Epa) of À2.8 . . . À3.0 V vs SCE. This value can be taken as lower
0
À
limit forEðR SiCl=R SiꢀþCl Þ which is expected not to be more negative.
3
3
Then, considering the shape of the log(k1)-E0
plot (in an ideal
P/Q
case when
a= 0.5) can give another idea of the cathodic limit of
0
À
EðR SiCl=R SiꢀþCl Þ. The intercepts of the activation-controlled portion
3
3
of the log(k1)-E0
plot (Fig. 3) with two adjacent diffusion-
P/Q
controlled portions (with 0 and 1/58 mV slopes) are equidistant
0
0
0
À
À
relative to EðR SiCl=R SiꢀþCl Þ ðat EðR SiCl=R SiꢀþCl Þ À EP=Q¼0Þ. Such sym-
3
3
3
3
metry of the activation zone also implies that
D
E0/0.058 < 2(log
(kD)–log(ks)) [24]. Therefore, from the half-spans of the visible
activation zones, the values of E0ðR SiCl=R SiꢀþCl of the chlorosilanes
ÀÞ
3
3
Fig. 3. Log(k1ap) versus Eo of redox mediators for the reaction of their anion radicals
in question cannot be more negative than À2.13, À2.25 and À2.10 V
Q with triorgano chlorosilanes and the estimation of E0
). With Me3SiCl
À
(SiCl/Siꢀ+Cl
for Me3SiCl, Et3SiCl and t-BuMe2SiCl, respectively (Fig. 3).
(&), data from [26] are shown. (ꢀ) Et3SiCl. For comparison, known values of k1 for
ap
The contribution of pure ET into k1 was estimated through
Ph GeCl (ꢃ) and Ph GeBr ( ) [30] are plotted.
4
3
3
fitting the experimental points with Marcus curve [26] defining log
(k1) as
With the concentration of [R3SiQꢀ] expressed as
!
2
ꢀ
ꢁ
F
D
G0
¼
logðk1Þ ¼ logZ À
D
G0 1 þ
(21)
d½R3SiQŠ=dt ¼ ½QŠðk2½R3Si ClŠ À k5½R3SiQŠ ¼ 0
(18)
(19)
(20)
¼
2:3RT
4DG0
with homogeneous collision frequency Z estimated (6.7 Â1010
MÀ1 sÀ1 for Me3SiCl) from Debye-Smoluchovski equation [50], and
½R3SiQŠ ¼ k2½R3SiClŠ=k5
¼
and with the ratio [R3Siꢀ]/[R3SiQꢀ] = k1/k2, one obtains
D
G0
=
l
/4 = 0.125 Â (lP
+ lSiCl). For aromatic anion radicals, sol-
vent reorganization energies lP have close values that can be
taken, following Kojima and Bard [51], as 0.63 eV. For chlorosilanes,
Àd½QŠ=dt ¼ ð2k1 þ 2k2Þ½QŠ½R3SiClŠ ¼ kap½QŠ½R3SiClŠ
1
the term lSiCl can be estimated as lSiCl = DSi-Cl + l0 [26]. It has been
previously shown [7,52] that in solution the Si—Cl bond undergoes
substantial polarization preceding the electroreduction of chlor-
osilanes. If the gas phase bond strength DSi-Cl (5.49 eV [53])
sustained in the solution, chlorosilanes would not be reducible
before -6 V vs SCE. So a good estimation of the effective Si-Cl bond
The corollary of this model is that k5 = k3 + k4, i.e. the rate of
consumption (SET, Eq. (12) of the primary radical product of the SN
step (Eq. (11), neglecting its dimerization under the conditions of
voltammetry, equals the sum of the rates of disappearance of silyl
radicals (k3 + k4) produced in the step 5a (k1). Both electron
transfers—one-electron (ET) and electron-pair (SN)—are reflected
in the kinetic law through the apparent rate constant6 (k1ap = 2
energy in solution, DSi-Clsol, is needed for assessing lSiCl and then
¼
D
G0 . First, it can be obtained according to the approximate
Eq. (22) [54]
k1 + 2k2). Therefore, considering the mediator P/Q couple (through
d
ip(P)) in terms of the
l
–ipcat/ip analysis of redox catalytic
sol
SiÀCl
D
ffi À2=3 Â ðEp À E0Cl=ClÀ Þ þ C
(22)
mechanisms [23], the response of this system will be complex
including the contributions form the two pathways. At this point, it
is not possible to distinguish these contributions since they have
similar reactants' energy profile and similar pre-exponential
factors; their competition intervenes through the interplay of
which, with Ep = À 3.1 V, and taking after Savéant E0
ꢀ
Cl /
À = 1.89 V and C = 0.3 eV (for C- and N-haloderivatives [54–56]),
Cl
provides DMe Si
Using more recent value E0
ffi 70:5 kcal molÀ1ð3:06 eVÞ as first estimate.
-
Clsol
3
À = 2.12 V [57] and C = 0.27 eV that
ꢀ
Cl /Cl
the corresponding activation energies in k1 and k2 in the apparent
ap
was directly calculated for the chlorosilane ðas C ¼
rate constant k1
.
-
2fðRT=
a
FÞInðZhetðRT=
a
F
n
DÞÞ - 0:78ðRT=
a
FÞ - 0=2 - TðS
l
Now supposing for the process a catalytic EC mechanism with
Me3SiCl
ap
SMe Si SCl Þg=3 [56] for v = 0.1 V sÀ1 with heterogeneous collision
ꢀ
an additional solution electron transfer (SET) (Eq. (9)) while k1
ꢀ
-
3
remains rate limiting step and using the working curve CAT–log(
(Fig. 1 from [23]; with = k1[P] and D are diffusion
2/D, where
layer thickness and the diffusion coefficient of P) with catalytic
l)
frequency Zhet = (kT/2
p
M)1/2 [58], diffusion coefficient D form
l
d
d
Stokes equation and the corresponding entropies obtained by
DFT calculations at B3LYP/Lanl2DZ level), one obtains
d
d
efficiency taken as CAT = 2(ipcat–ipd)/ip SiCl (here ipcat, ipd and ip
DMe SiÀClsol ffi 74:04 kcal molÀ1ð3:21 eVÞ . Provided that bond
SiCl
3
are peak currents of P in the presence and the absence of R3SiCl and
the peak current of R3SiCl, respectively) provided the values of log
strength is enthalpy, both these approaches have to treat the
entropic contribution to energy, included in term C, either
supposing it to be small and practically constant [54,55] or
(k1ap), plotted as a function of E0
in Fig. 3.
P/Q
Reduction of Me3SiCl, Et3SiCl and t-BuMe2SiCl with the
explicitly calculating it. A more complete way for estimation of DSi-
mediators listed in Table 1 corresponds to an activation controlled
sol
process with the
112 mV and 1/108 mV, respectively. For none of these chlorosilanes
D
log(k1)/
D
E0
slopes of À1/105 mV [36], 1/
(DGSi-Cl) can be calculating the profile of Gibbs energy of the
Cl
P/Q
reaction via frequency analysis (here at DFT B3LYP/Lanl2DZ level)
when solvation is taken into account using an appropriate model
(e.g. [59]). For Me3Si-Cl, thus obtained free enthalpy of Si—Cl bond
dissociation in DMF amounts to 2.048 eV (see Section 3.2).
6
Á
Similar expression would result in the extreme case if the reduction of R3Si
(Eq. (6)) entirely occurs at the cathode, via heterogeneous ET (i.e. when k3 = 0).

462
S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
Table 2
Parameters of redox-mediated reduction of alkyl chlorosilanes in DMF/0.1 M Bu4NBF4
Ep, Va
l0
DSi-Cl
k1 (ET)c
D
G
(ET)b
k2 (SN)c
b
solb
¼
¼
DG
(SN)b
E0ðR
Va
3 SiCl=R3 SiꢀþClÀÞ
R3SiCl
Me3SiCl
À3.10
À2.87
À2.82
À2.06 (À1.89)
À2.11 (À1.94)
À2.00 (À1.83)
42.36 (1.836)d
25.16 (1.091)
42.49 (1.842)
26.34 (1.142)
42.58 (1.846)
42.18 (1.829)
74.04 (3.210)
47.23 (2.048)
70.48 (3.056)
47.99 (2.081)
69.72 (3.023)
49.82 (2.160)
0.053
19.18 (0.832)
13.61 (0.590)
18.51 (0.803)
13.20 (0.572)
18.39 (0.798)
16.20 (0.703)
4.73 Â104
3.91 Â103
2.95 Â104
4.59 Â104
3.61 Â102
3.53 Â102
11.07 (0.480)
12.55 (0.544)
11.35 (0.492)
11.09 (0.481)
13.96 (0.605)
13.97 (0.606)
6.43 Â102
0.165
Et3SiCl
1.29 Â103
t-BuMe2SiCl
0.201
8.16
a
V vs. SCE; in parentheses, the E0 corrected for liquid-liquid potential for DMF/water, 0.172V [57].
b
c
kcal molÀ1 (eV).
MÀ1 sÀ1
d
Upper row values: derived with l0 from Born model [26] and DSi-Cl from Eq. (22); lower row: from DFT B3LYP/Lanl2DZ calculation of reaction
DG profiles.
Similarly obtained values for other chlorosilanes seem consistent
and were used in the following (Table 2).
it is possible to separate them from Fig. 4 (k1ap - 2 k1 = 2k2), using
the calculated k1 for pure ET case, and to determine the rate of the
Standard potentials E0 of ET to alkylchlorosilanes,
nucleophilic attack of the anion radical of any given mediator on
¼
0
¼
the chlorosilane (Table 1). It is to note that with the
DG0
À
EðR SiCl=R SiꢀþCl Þ, necessary for
D
G0 determination, were obtained
3
3
estimated form the approximate model (Eq. (22), the k2 values are
superior to k1 by the factor of ca. 105-106 (e.g. for Me3SiCl/
naphthalene, k1 = 1.98 Â 102 and k2 = 6.27 Â107 MÀ1 sÀ1), which
according to the thermochemical scheme [60] which proved
efficient for calculating E0 of alkyl halides [61]. Here, given the lack
of consistent data on free enthalpies of formation of silyl radicals
seems greatly underestimating the contribution of ET. Using
DGSi-
D
Gf0(R3Siꢀ) [53], the necessary energies of formation have been
from DFT calculations (see 3.2) provides better agreement with
Cl
calculated for all entities at DFT B3LYP/Lanl2DZ level at T = 298.15 K
and P = 1 bar (Ox = R3SiCl, Red = R3Siꢀ + ClÀ). Solvation has been
taken into account using Tomasi's isodensity polarized continuum
model (IPCM) for model reaction field and analytic DFT energies
calculation [59]. Obtained from DFT calculations, the values of E0
are not related to any reference electrode, so they were brought to
the SCE scale (ESCE = 0.241 V vs. NHE), using recent value of absolute
the experiment. Thermodynamic parameters of the SN process can
then be assessed from the values of k2 using a retro-procedure
through the transition state theory equation for the rate constant
¼
of a bimolecular reaction, k2 =
k
P(kT/h) exp(À
D
G /RT) [63]. Here,
the values of the transmission coefficient
k and steric factor P (as
Sster /R), [64]) are not known; in first approach they are
¼
exp(
D
assumed to be equal to unity for Me3SiCl and Et3SiCl though they
potential of standard hydrogen electrode proposed by Gennaro,
¼
are supposedly smaller for t-BuMe2SiCl. Thus estimated
DG
+
E0 /H2 = 4.281 V [57]. The influence of ion pairing with Bu4N+ was
H
values correspond to quite low activation barriers (Table 2), which
is consistent with high reactivity of chlorosilanes in SN-reactions.
supposed to be small, see [62].
0
À
EðR SiCl=R SiþCl Þvs:SCE ¼
D
G0f ðOxÞðgÞ
À
D
G0f ðRedÞðgÞ
þ
ð 3
D
G0ðR Si ClÞ
À
D
G0ðR SiÞsolv À
D
G0ðClÀ solv Þ À 4:522
3
(23)
3.2. DFT calculations
solv
3
Þ
3
Consistency of the electrochemical results was checked
considering theoretical energy profiles of the corresponding
processes. Potential energy (PE) and free Gibbs energy profiles
for an ET-induced reductive cleavage of Si-Cl bond in Me3SiCl have
been calculated at DFT B3LYP/Lanl2DZ level in gas phase and in
DMF (IPCM calculations, solvent cavity built using scaled van der
Waals surface). PE profiles, plotted at zero driving force, are shown
in Fig. 5. Remarkably, the PE profile for (Me3Siꢀ + ClÀ) system in the
gas phase shows a shallow minimum which can be related to a
À0.202 eV stabilized Me3SiClꢀ intermediate with ꢄ20% stretched
Si—Cl bond (Fig. 5a). In DMF, solvation of the fragments overruns
this stabilization and only repulsive part appears; the situation
0
Now with DSi-Clsol and EðR SiCl=R SiꢀþCl Þ in hands, the correspond-
À
3
3
¼
ing standard activation energy
D
G0 could be determined and
hence, from Eq. (21), the theoretical k1 (Table 2). Plotted versus E0
P/
Q, the k1 values (dotted lines in Fig. 4) thus correspond to the pure
ET scheme. As is seen for Me3SiCl, a fair fit of the experimental
points with Marcus parabola, assuming pure ET (Fig. 4), was found
with Z = 6.7 Â1010 and Eð0Me SiCl=Me SiꢀþCl ¼ À2:06V for
D
G0 = 0.34
¼
ÀÞ
3
3
¼
eV. Similar trends with close values of
and t-BuMe2SiCl. Obviously, such standard activation energy
appears too small for process including bond-breaking.
DG0 were found for Et3SiCl
a
Therefore,eitherthereductionofthechlorosilanesisnotdissociative
in its nature and involves transient anion radicals (though not
detectable by voltammetry), as was supposed for the reduction of
ClCH2TMS[36],orthatbondbreakinginchlorosilanesrequiresinfact
much lower energy than estimated above. Following the model of
parallel ETand SN processes, it might also mean that the whole set of
experimental log(k1ap) values is uplifted relative to pure ET process
(k1) due to the contribution of parallel nucleophilic interaction of Q
with R3SiCl. Taking into account possible partial contamination of
the catalyticcurrent with somecontributionfromEq. (12) would not
explain such a large gap since this contribution is expected to be
smaller than that of SN. It is to be noted that similar excessive log(k1)
values were observed for homogeneous reduction of n-BuI
compared to sec- and tert-BuI [61] when a substantial contribution
from an SN2 process was supposed.
thus corresponds to
a pure dissociative ET (Fig. 5b). For
t-BuMe2SiCl (Fig. 6a), such a minimum is less pronounced and
is shifted to longer lSi-Cl, corresponding to long-term coordination
interactions between t-BuMe2Siꢀ radical and ClÀ.
For both chlorosilanes, Si-Cl bond dramatically weakens when
going form the gas phase to solution (Figs. 5 and 6). Solution bond
energies from the PE profiles seem more appropriate for
¼
calculating
D
G0 than those obtained using the approximate
model (Eq. (22). Yet even better way of assessing Si—Cl energies
and intrinsic barriers is to consider pure Gibbs energies (Fig. 5c)
available from frequency analysis of the optimized configurations
along the PE profile. Using these free energies of bond dissociation
in DMF (Table 2), theoretical ET rate constants (k1) are closer to the
ap
experimentally obtained k1 (Fig. 4). Now realizing the risk of
direct comparison of the results of voltammetry and of large-scale
electrolysis, a rough approximate of the ratio k1/k2 can be obtained
from the practical ratio of ET to SN products in large scale
As mentioned above, left sides of Eqs. (5a) and (5b) are similar
but these reactions have different transition states. Now provided
the rate constants k1 and k2 are included into k1ap through Eq. (20),

S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
463
reductions, which is roughly ranging from ꢄ1:10 to ꢄ1:100 (e.g. for
Me3SiCl and Et3SiCl using anthracene anion radicals, Table 3). Then
the k1 are supposed to be of the order of k1ap/(20 . . . 200), which is
in a good accordance with the above theoretical estimates.
In DMF, which is more efficient silicophile than pyridine or even
Me3N [2], Si—Cl bond is affected not only by solvation but also by
nucleophilic assistance of the solvent. To take this interaction into
account, energy profile of the system Me3SiClꢀDMF along the ET
process has been calculated (Fig. 5d). Its striking feature is that the
minimum related to the ground state geometry (lSi-Cleq) is now only
slightly marked: all following geometries (for lSi-Cl > 2.25 Å) are
almost equivalent in energy. The curve of the products shows that
past equilibrium geometry, the energy decreases linearly with
distance until ꢅ3.7 A, after which it remains practically unchanged.
The feature is quite similar for t-BuMe2SiCl, though bond breaking
energy is more pronounced(Fig. 6c), probably because approaching
of DMF molecule is impeded by higher steric demand. Equilibrium
geometries for both chlorosilanes clearly correspond to penta-
coordinate species. Evidently, bond energies DSi-Cl in this situation
are not the same and are not suitable for using in the model
considered in 3.1.
Although electrochemistry of hyper-coordinate silicon species
is still to be developed, some idea on relative electrophilicity of
silicon in DMF-coordinated Me3SiCl with respect to its non-
coordinated form can be obtained from NPA charges and from the
relative LUMO levels in these species (Scheme 1).
In spite of smaller partial positive charge on Si (1.608 vs. 1.721),
LUMO level in DMF-associated penta-coordinated chlorosilane is
0.408 eV lower than in the non-associated form rendering the
electron uptake easier. In Cl-silatrane, reduced at quite similar
potentials and having strong intramolecular N ! Si coordination
[65], the charge at Si is remarkably more positive (2.228) while
LUMO energy is only 0.1 eV higher than in Me3SiClꢀDMF. It seems
that electron uptake in these systems is mainly orbital-controlled.
Thusdonorinteractionsfromanexternaloraninternalsilicophile
do not increase LUMO energy in such species and therefore do not
make their reduction more difficult, contrary to a common sense
inspired by carbon chemistry. This additional energy is used for
rearrangement of the orbitals of Si upon coordination expansion,
generally lowering LUMO energy, promoting nucleophilicity and
rendering the reduction easier. Evidently, careful quantification of
nucleophilicassistance ofsilicophilesisneededfor the development
of a more elaborated model of reactivity of chlorosilanes.
3.3. Large scale reductions
3.3.1. Anthracene
When reducing anthracene in the presence of sub-equivalent
amounts of Me3SiCl, a kinetic pre-peak (Epp) with
DE = Ep–
Epp ffi 300 mV appears on the voltammogram, caused by a fast
nucleophilic attack of anthracene anion radicals on the chlor-
osilane. Very electrophilic character of Me3SiCl renders its
reactions with even poorly nucleophilic delocalized anion radicals
quite efficient. Indeed, with E0P/Q – Ep of 1.21 V, the estimated ratio
of the rate constants of homogeneous ET to SN between Q and
Me3SiCl (Table 1) can be compared with the reported alkylation of
aromatic anion-radicals in a formal SN2 process [26]. In the case of
alkyl halides, the contribution from SN2 was essentially neglected
compared to ET mechanism [66]. For alkyl chlorosilanes, and to the
greatest extend for Me3SiCl, the situation is inversed: SN-type
process is systematically favored compared to ET.
Fig. 4. Quadratic fits for the experimental k1ap (solid line) and Marcus plots for pure
ET (dashed line: DSi-Cl from Eq. (22); dotted line: GSi-Cl from DFT calculations) for
D
homogeneous reduction of three trialkyl chlorosilanes with aromatic anion
radicals.
(24)

464
S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
Fig. 5. Reduction of Me3SiCl. (A) PE profile in the gas phase, (B) PE in DMF (polarized continuum with e= 36.7). (C) Free Gibbs energy and DS (right scale, &) in DMF. (D) PE
profiles for the reduction in the presence of one molecule of DMF in DMF; the arrow corresponds to the structure shown. From DFT B3LYP/Lanl2DZ.
Fig. 6. PE profiles for the reduction of t-BuMe2SiCl: (A) in the gas phase, (B) in DMF. (C) Reduction in the presence of one molecule of DMF (in DMF as solvent). From DFT
B3LYP/Lanl2DZ.
3.3.2. Naphthalene
With naphthalene, the contribution of ET process amounts to
ꢄ8%, according to the amount of formed disilane. Instead, the
(25)
silylation of naphthalene was quite efficient. Interestingly, two
disilylated isomers - 1,4-bis(triethylsilyl)-1,4-dihydronaphthalene
and 1,2-di(triethylsilyl)-1,2-dihydronaphthalene - were formed
(Table 1), arising from two radical anion forms of naphthalene
(Eq. (27)). The ratio of these products is 1:1.7, the major isomer being
supposedly1,2-silylatedone.Theseresultsagreewiththoseobserved
for the reaction of chemically prepared naphthyl lithium with
(26)
Me3SiCl[68] andMe3SiCN[69].Interestingly,therecoveryofstarting
naphthalene was reported in the chemical version of this process,
presumingthattheNaphꢀ alsointervenesaselectrontransferreagent
providing an electron for the reduction of intermediate monosily-
lated naphthyl radical to the corresponding anion [68].
In the large scale reduction of the anthracene/chlorosilane
mixture, only traces of the disilane (ET product) were observed by
GC-MS. Instead, 9- and 9,10-silylated anthracenes were formed as
main products (Table 3). Silyl-substitution at the benzylic position
9 in anthracene does not affect much electronic demand of the
primary silylated radical but it neutralizes the negative charge of
this species. Since neutral radicals are usually easier to reduce than
the starting neutral molecule (|E2| < |E1|), once nucleophilic attack
(Eq. (25)) produces silylated dihydroanthryl radical, it undergoes
immediate reduction to form the corresponding anion. The mono-
silylated anion follows second silylation at the position 10. Small
amount of di-9,9'-silylated dimer of anthracene was also formed,
arising from 10,10'-dimerization of 9-substituted intermediate
radicals, in a way similar to that reported by Hammerich and
Parker [67].
(27)

Table 3
Electroreduction of aromatic mediators (P) in the presence of trialkyl chlorosilanes in THF-DMF (10:1 v/v)/0.1 M Bu4NBF4
Chlorosilane
Me3SiCl
Mediatora Rb Bc, % Product
AN 3.0 traces
m/z, %
Yield, %d
73.2
252 (M+, 2), 237 (M-Me, 98), 179 (M-Me3Si, 61), 91 (C7H7+, 21), 77 (Ph), 73 (Me3Si+, 31), 65 (C5H5+, 7), 51
324 (M+, 1), 309 (M-Me, 100), 294 (-2Me, 21), 251 (M-Me3Si, 34), 91 (C7H7+, 37), 77 (Ph), 73 (Me3Si+, 24), 65 (C5H5+), 51
22.6
Et3SiCl
AN
5.6 9.6
294 (M+, 8), 279 (M-Me, 95), 265 (M-Et, 62), 179 (M-Et3Si, 76), 115 (Et3Si, 36), 65 (C5H5+, 5), 51, 43
56.5
27.8
408 (M+), 379 (M-Et, 5), 350 (M-2Et, 27), 321 (M-3Et3, 1), 293 (M-Et3Si, 24), 115 (Et3Si+, 100), 91, 77, 65, 51
293 (M/2, 19), 279 (M-Me, 71), 265 (M-Et, 100), 179 (M-Et3Si, 52), 115 (Et3Si, 47), 65 (C5H5+), 51 (C4H3+, 12), 43
< 5
Et3SiCl
NP
3.5 8.2
244 (M+, 2), 229 (M-Me, 15), 215 (M-Et, 27), 186 (M-2Et, 2), 157 (M-3Et, 2), 115 (Et3Si, 34), 104, 91 (C7H7+, 23), 77 (Ph, 11), 65 (C5H5+), 51 47.6
358 (M+, 8), 343 (M-Me, 32), 329 (M-Et, 11), 243 (M-Et3Si, 15), 115 (SiEt3+, 100), 91, 77, 51 (C4H3+, 8), 43
358 (M+, 0.5), 343 (M-Me, 56), 329 (M-Et, 0.5), 301 (M-2Et, 12), 243 (M-Et3Si, 6), 115 (SiEt3+, 100), 91, 77, 51
23.3
13.2
486 (M+, 2), 457 (M-Et, 9), 243 (M/2, 23), 215 (M/2–2Et, 8), 115 (SiEt3, 100), 91, 77, 65
6.8

Table 3 (Continued)
Chlorosilane
Mediatora Rb Bc, % Product
m/z, %
Yield, %d
Et3SiCl
MBA
3.5 16.2
252 (M+, 32), 237 (M -Me, 9), 221 (M-OMe, 11), 115 (Et3Si, 98), 122 (C7H6O2+, 13), 105 (C7H5O+, 19), 77 (Ph, 25), 65, 51, 32 (CH4O, 9)
366 (M+, 94), 337 (M -Et, 5), 309 (M-2Et, 8), 251 (M-Et3Si, 23), 123 (C7H7O2+, 21), 115 (Et3Si, 100), 105 (C7H5O+, 8), 77 (Ph, 19), 65, 51
78.1
2.8
Et3SiCl
BN
2.4 20.5
333 (M+, 7), 304 (M-Et, 100), 275 (M-2Et, 18), 218 (M-Et3Si, 47), 115 (Et3Si, 31), 102 (12), 77, 51
52.2
e
306 (M+, 3), 277 (M-Et, 100), 249 (M-2Et, 25), 191 (M-Et3Si, 32), 115 (Et3Si+, 56)
3.2
7.4
5.3
192 (M+, 5), 163 (M-Et, 100), 135 (M-2Et, 9), 115 (Et3Si+, 43), 105 (C7H5O+, 6), 77, 42
268 (M+, 3), 253 (M-Me, 72), 239 (M-Et 93), 191 (M-Ph 48), 153 (M-SiEt3, 61), 115 (Et3Si, 100), 77 (Ph, 34), 65, 51
t-BuMe2SiCl
AN
3.5
–
t-BuMe2SiH
115 ([M-H]+, 76), 100 (M-Me, 47), 65 (100), 58 (M-t-Bu, 15), 57 (t-Bu, 43), 43 (MeSi, 10)
53.5
34.3
296 (M+, 4), 281 (M-Me, 78), 115 (Et3Si+, 24), 181 (M-Et3Si, 100), 91, 77 (Ph, 7), 65 (C5H5+, 13), 57 (t-Bu, 11)
a
AN: anthracene, MBA: methylbenzoate, NP: naphthalene, BN: benzonitrile.
Ratio [Et3SiCl]/[P].
b
c
Yield of the corresponding disilane (Me3SiSiMe3,m/z: 146 (M+, 17), 131 (M-Me, 38), 115 (M-2Me, 7), 73 (Me3Si, 100); Et3SiSiEt3, m/z: 230 (M+, 15), 201 (M-Et, 16), 115 (Et3Si+, 77)).
d
Yields are based on the reacted R3SiCl. Unreacted R3SiCl was converted into corresponding disiloxane during the work-up procedure (Me3SiOSiMe3, m/z: 147 (M-Me, 100), 131 (M-2Me, 7), 117 (M-3Me, 1) 73 (Me3Si, 15);
Et3SiOSiEt3, m/z: 246 (M+, 2), 217 (M-Et, 100), 188 (M-2Et, 79), 115 (Et3Si, 37)).
e
Supposed structure; 1H NMR (CDCl3, TMS): 7.78 (3H), 7.16 (2H), 1.10 (18H, CH3), 0.63 (12H, CH2).

S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
467
-0.516
-0.392
-0.513
-0.460 Cl
-0.211
Cl1.608
-0.430 Cl
(31)
Si 1.019
1.721
O
Si 1.113
O
2.228Si
O
1.432 N
-1.141
-0.981
-0.730
O
-0.380
-1.170
Protonation and silylation of the carbanion resulting from the
reduction of the intermediate silylated radical (Eq. (31)) lead to the
main product, corresponding silyl(methyl) acetal. Contribution of
pure ET from this mediator to Et3SiCl, from the yield of Et3SiSiEt3, is
16% only.
-0.954
-0.663
-0.409
N
Scheme 1. NPA (bold) and Mulliken (italic) charges in Me3SiCl, Me3SiClꢀDMF and
chlorosilatrane in DMF from DFT B3LYP/Lanl2DZ calculations.
3.3.5. Bipyridine
The anion radicals of 2,2'-bipyridine are
p-delocalized akin to
For comparison, Li-reduction of PhH in the presence of Me3SiCl
yielded 1,4-bis(trimethylsilyl)-1,4-dihydrobenzene [70] in a “silyl
Birch reduction” of aromatics”.
those of the above aromatic mediators, so they were expected to
trigger similar reactions with chlorosilanes. However, 2,2'-
bipyridine has lone pairs at N atoms, perpendicular to its
p
-system, so that it can manifest its silicophilic character both
3.3.3. Benzonitrile
in the neutral form and in the anion radical. Being better silicophile
than DMF [1], bipyridine is able to pre-coordinate with chlor-
osilanes before ET, thus making the whole reaction scheme
different. Bipyridine is therefore falling out the reaction series of
aromatic mediators (sf. Table 1), though the reduction of
chlorosilanes in its presence also results in Si—Cl bond cleavage.
Upon progressive addition of R3SiCl to 2,2'-bipyridine, initially
showing a reversible redox system, the anodic counterpart of the
reduction of bipyridine (at Ep = À2.21 V) progressively vanished but
the reduction peak did not show any catalytic increase. At the same
time, this system exhibited a new reduction peak at about
Although providing a considerable driving force for the ET
(Ep–E0P/Q, Table 1) comparable to that of methylbenzoate, and
being most efficient as ET reagents in the reaction series
considered (Table 3), the anion radicals of benzonitrile have quite
complex own reactivity.
The elimination of CNÀ anion might generate Phꢀ radical whose
reactions could explain the formation of the silylated biphenyl,
PhC6H4SiEt3. This product might arise from radical phenylation of
PhSiEt3, formed through Eq. (28), or from nucleophilic substitution
ꢀ
of (C6H4)2 anion radical on Et3SiCl. Due to small electron-
withdrawing effect (sp = 0.02 [71]), Et3Si-group cannot lower the
reduction potential of PhSiEt3 to ensure its efficient separate
reduction; in addition, electroreduction of the homologous
PhSiMe3 was reported to result in silylated cyclohexa-1,4-dienes
[72]; such products were not formed in the electrolysis.
In neutral benzonitrile, charge distribution is such that carbon
of the CN-group is most susceptible to nucleophilic attacks, but in
the reduced forms of benzonitrile the negative charge is mostly
localized on N atom according to ROHF/6–311 G calculations. This
features account for the formation of the product of addition of
Et3SiÀ anion across the CN bond followed by N-silylation of the
resulting anion (Eq. (29)); this product (with m/z = 333) was
tentatively characterized as N-silylated ketimine (sf. [73]).
E
p' = À1.1 V, the exact value depending on the chlorosilane taken.
The Ebipy - Ep' for this new peak is too large to be provoked by the
kinetic shift of the reduction of bipyridine, so it supposedly
corresponds to the reduction of a complex between 2,2'-bipyridine
and the chlorosilane. Such complexes have been prepared and
isolated in case of triorganobromo- and iodo- silanes and
polychlorosilanes [76], but not for monochlorosilanes. However,
these complexes might exist in the solution, though with a small
stability constant, and their formation is well seen through their
voltammetric response (Fig. 7). Here, the peak p1 (Ep = À1.1 V)
arises from the reduction of Et3SiClꢀbipy complex and the peak p2
is the oxidation of ClÀ anion eliminated at the step p1. The latter
fact suggests the occurrence of intramolecular electron transfer
between the two units: from SOMO (
ligand anion radical to LUMO of the chlorosilane (Si-Cl
The formation of complexes of electrophilic silicon halides with
p
-system of bipy) of the
s
*-bond).
(28)
(29)
3.3.4. Methylbenzoate
Supposedly, because of large affinity of Si for O favoring the
formation of O-silylated products, no phenyl silylation was
detected, as in the reaction of benzophenone anion radical with
2-bromooctane [74]. This is consistent with the fact that O atom in
carbonyl anion radicals acts as a nucleophilic center while C-atom
mostly intervenes in ET processes [75].
(30)
Fig. 7. Voltammograms of reduction of 2,2'-bipyridine (3 Â10À2 mol LÀ1
) in
CH3CN + 0.1 M Et4NBF4 at a GC electrode: (a)—before and (b)—after addition of
Et3SiCl (2.4 Â10À2 mol LÀ1). Temperature 21 ꢁC. v = 500 mV sÀ1. With DMF used as
solvent, electrode passivation occurred at [Et3SiCl] > 3 Â10À3 mol LÀ1
.

468
S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
2,2'-bipy and other N-bases is now well documented [77] and
electroreduction of homologous complexes of germanium chlor-
ides with 2,2'-bipyridine has been recently reported [78].
In terms of the reduction potential, the complexation of a
chlorosilane with bipyridine formally corresponds to introduction
of an acceptor substituent into the bipy moiety. Its equivalent
model is not simple and needs a special consideration. Coordina-
tion with the silicophilic solvent (DMF) is effective throughout the
whole process. Its contribution supposedly evolves when going
form one silylated intermediate to another but at the present state,
it is impossible to provide
a quantitative account for this
phenomenon. Given a large excess of the silicophile at any
moment, one can suppose that pseudo-first order conditions for
these interactions were maintained, and hence, in a first approach,
they could be neglected as supposedly constant.
electron-withdrawing ability was estimated through E1/2
correlations from the reaction constants (
–rs
r
= 0.9–1.1) derived from
the reduction E1/2 of 4-R-2,20-bipyridines and the ligand-based
reduction of their RuL3 complexes [79], and ligand-based
reduction potentials of Re complexes with substituted bipyridines
fac-[4,40-X2–2,20-bipy]Re(CO)3Etpy+ [80]. Thus obtained plot
Electrochemical silylation of delocalized aromatic anion
radicals is not so chemoselective as in the case of
s-anions or of
non-delocalized alkyl carbanions; charge delocalization and
radical coupling account for the formation of multiple silylated
products. Interestingly, that in case of t-BuMe2SiCl, most favorable
for observing ET chlorosilane, the t-BuMe2Si-group is too bulky for
an efficient SN self-process of the silyl anion with the starting
chlorosilane; under these conditions, no disilane was formed and
t-BuMe2SiÀ anion ends up as an electrochemically inactive
hydrosilane t-BuMe2SiH.
In general, the present study has shown that the use of redox
mediators, in spite of its expected advantage as compared to direct
cathodic reduction of chlorosilanes, is quite inefficient for
preparing disilane products. On the contrary, silylation of the
aromatic anion radicals via an SN–like process seems to be a good
complement to the conventional synthetic methods involving
aromatic reductions with alkali metals but being free of the
inconveniences of the latter. Electroactive silicophilic ligand 2,2'-
bipyridine forms electroactive hyper-coordinated species whose
low-potential cleavage of Si—Cl bonds merits a special study.
Further works in this direction are under the progress.
(Ep = rsp–2.1) allowed assessing the
s-constant of an equivalent
substituent at p-position of bipyridine that would provoke similar
anodic shift in Ep of bipy as that due to the complexation with
R3SiCl. The obtained value, sp = 1.8 for a single 4-R substituent (or
sp = 0.9 for each of two 4,40-R2- substituents), is equivalent to the
electron-acceptor effect of p-(CF3)3CSO2 group or of two p-Me2 S+
groups [71]. Rough estimation of the complex formation constant,
Kf, from E0(bipy) and E0(bipyꢀR3SiCl), assuming the reaction
between bipyꢀ and R3SiCl to be total with K > 104, provides
Kf < 8 Â 1010. The study of these systems is presently in progress.
4. Conclusions
In a series of representative alkyl chlorosilanes, the rates of their
reactions with electrogenerated aromatic anion radicals were
determined by homogeneous redox catalysis method. Thus found
ap
apparent rate constants k1 are remarkably higher than those of
pure one-electron transfer to alkyl chlorosilanes expected
according to Marcus model. The difference between these
constants arises from the contribution of the parallel two-electron
(nucleophilic) interaction between the mediators and chlorosi-
lanes. The rates of nucleophilic displacement at Si by aromatic
anion radicals, estimated from the comparison of the experimental
log(k1ap) and theoretical log(k1)–EP/Q plot, are quite high amount-
ing to 102-108 MÀ1 sÀ1. These values are comparable with those
found for the total reaction (including both ET and SN contribu-
tions) between aromatic anion radicals and hexyl iodide [34] and
are higher than those determined for n-BuI after subtraction of the
ET term [61], which agrees with higher electrophilicity of
chlorosilanes compared to alkyl iodides.
For an SN-like process, the steric demand determining the
availability of Si becomes more important when going from
Me3SiCl to t-BuMe2SiCl, so the rate of the nucleophilic reaction of
alkyl chlorosilanes with the same aromatic nucleophile decreases
in average by 4–9 times form Me3SiCl to Et3SiCl but may drop by
about 200–300 when passing form Et3Si to t-BuMe2Si derivatives.
Steric hindrance brought by t-Bu group in t-BuMe2SiCl affects the
Acknowledgements
The authors wish to thank the support from the University of
Setif (travel grant for SS) and the University of Dakar (travel grant
for MD). Financial aid from Rennes Metropôle (acquisition of the
GC–MS) is gratefully acknowledged.
References
[1] M.A. Brook, Silicon in Organic, Organometallic and Polymer Chemistry, John
Wiley & Sons, Toronto, 2000, pp. 52.
[2] A.R. Basindale, P.G. Taylor, Reaction Mechanisms of Nucleophilic Attack at
Silicon, in: S. Patai, Z. Rappoport (Eds.), The Chemistry of Organic Silicon
Compounds, John Wiley & Sons, 1989, pp. 840.
[3] J.L. Brefort, R.J.P. Corriu, C. Guerin, B.J.L. Henner, W.W.C. Wong Chi Man,
Organometallics 9 (1990) 2080.
[4] R.J.P. Corriu, C. Guerin, J.J.E. Moreau, The Chemistry of Organic Silicon
Compounds, in: S. Patai, Z. Rappoport (Eds.), John Wiley & Sons, 1989, pp. 334.
[5] R.J.P. Corriu, G. Dabosi, M. Martineau, J. Chem. Soc. Chem. Commun. 10 (1979)
457.
[6] V. Jouikov, Curr. Topics Electrochem. 4 (1997) 1.
[7] V.V. Zhuikov, Rus. J. Electrochem. 36 (2000) 117.
[8] X. Wang, Y. Yuan, I. Cabasso, J. Electrochem. Soc. 152 (2005) E259.
S
N process to much lesser extent compared to t-Bu effect in carbon
chemistry. This feature stems from longer Si—C bonds compared to
C—C, generally providing poorer shielding of Si by alkyl sub-
stituents. Though this order is in good agreement with the general
reactivity of electrophilic trialkylsilyl derivatives Me3Si > Et3Si >> t-
BuMe2Si [1–4,81], steric demand of the nucleophile should also be
taken into account for a more strict consideration. Corresponding
activation energies agree with the observed contributions from ET
and SN pathways in all three reaction series. The approximate and
the DFT-supported determinations of thermodynamic parameters
for the ET model allowed assessing the Si-Cl bond strength in
solution, which becomes more than 3 eV weaker compared to the
gas phase because of polarization(/coordination) by(/with) the
solvent and the interactions with the supporting electrolyte cation
[6,52]. However, even if these two phenomena evidently account
for a part of the reorganization energy and therefore influence the
outcome of the formal Marcus treatment, their inclusion in the
[9] V. Jouikov, Rus. Chem. Rev. 66 (1997) 509.
[10] W.G. Boberski, A.L. Allred, J. Organomet. Chem. 88 (1975) 73.
[11] E. Hengge, G. Litscher, Angew. Chem. 88 (1976) 414.
[12] R.J.P. Corriu, G. Dabosi, M. Martineau, J. Organomet. Chem. 188 (1980) 63.
[13] E. Hengge, H. Firgo, J. Organomet. Chem. 212 (1981) 155.
[14] A. Kunai, T. Kawakami, E. Toyoda, M. Ishikawa, Organometallics 10 (1991) 2001.
[15] M. Bordeau, C. Biran, M.P. Leger-Lambert, F. Sparau, D. Deffieux, Inorganic and
Organometallic Polymers II. Advanced Materials and Intermediates, in: P.
Wisian-Neilson, H.R. Allcock, K.J. Wynne (Eds.), Kluwer Academic Publishers.
ACS Symposium series, Washington, 1994, pp. 18.
[16] D.G. Peters, Organic Electrochemistry, in: H. Lund, O. Hammerich (Eds.), 4th
Ed., Marcel Dekker, NY, 2001, pp. 341.
[17] A.H. Holm, T. Brinck, K. Daasbjerg, J. Am. Chem. Soc. 127 (2005) 2677.
[18] J. Zeitouny, V. Jouikov, Phys. Chem. Chem. Phys. 11 (2009) 7161.
[19] A.J. Fry, Synthetic Organic Electrochemistry, Harper & Row, NY, 1972, pp. 67.
[20] J.W. Sease, R.C. Reed, Tetrahedron Lett. 6 (1975) 393.
[21] H. Lund, J. Simonet, J. Electroanal. Chem. Interfacial Electrochem. 65 (1975)
205.
[22] J. Simonet, J.F. Pilard, Organic Electrochemistry, in: H. Lund, O. Hammerich
(Eds.), 4th Ed., Marcel Dekker, NY, 2001.

S. Soualmi et al. / Electrochimica Acta 158 (2015) 457–469
469
[23] C.P. Andrieux, J.M. Dumas-Bouchiat, J.-M. Saveant, J. Electroanal. Chem. 113
(1980) 1.
W., Chen, M.W., Wong, C. Gonzalez, and J.A. Pople, Gaussian, Inc. Pittsburgh PA,
(2003).
[24] C.P. Andrieux, J.M. Dumas-Bouchiat, J.-M. Saveant, J. Electroanal. Chem. 87
[46] C.P. Andrieux, J.M. Saveant, Investigations of Rates and Mechanisms of
Reactions, in: C.F. Bernasconi (Ed.), Wiley, New York, 1986 (Chapter VII).
[47] C.P. Andrieux, I. Gallardo, J.-M. Saveant, K.B. Su, J. Am. Chem. Soc. 108 (1986)
638.
(1978) 39.
[25] C.P. Andrieux, C. Blocman, J.M. Dumas-Bouchiat, J.M. Saveant, J. Am. Chem. Soc.
101 (1979) 3431.
[26] J.M. Saveant, Elements of Molecular and Biomolecular Electrochemistry, John
Wiley & Sons, New Jersey, 2006.
[27] T. Lund, R. Fuhlendorff, D. Occhialini, S.U. Pedersen, H. Lund, Acta Chem. Scand.
43 (1989) 803.
[48] P.P. Gaspar, A.D. Haizlip, K.Y. Choo, J. Am. Chem. Soc. 94 (1972) 9032.
[49] C. Chatgilialoglu, J.C. Scaiano, K.U. Ingold, Organometallics 1 (1982) 466.
[50] C. Amatore, Organic Electrochemistry, in: H. Lund, O. Hammerich (Eds.), 4th
Ed., Marcel Dekker, NY, 2001, pp. 25.
[28] D. Occhialini, S.U. Pedersen, H. Lund, Acta Chem. Scand. 44 (1990) 715.
[29] D. Occhialini, J.S. Kristensen, K. Daasbjerg, H. Lund, Acta Chem. Scand. 46
(1992) 474.
[51] H. Kojima, A.J. Bard, J. Am. Chem. Soc. 97 (1975) 6317.
[52] V. Jouikov, C. Bernard, Nouv. J. Chem. 23 (1999) 287.
[53] R. Walsh, The Chemistry of Organic Silicon Compounds, in: S. Patai, Z.
Rappoport (Eds.), John Wiley & Sons Ltd., 1989.
[30] K. Daasbjerg, Acta Chem. Scand. 47 (1993) 398.
[31] A.V. Bukhtiarov, A.P. Tomilov, V.P. Bud'ko, Elektrokhimiya 16 (1980) 1274.
[32] J. Yoshida, K. Muraki, H. Funahashi, N. Kawabata, J. Organomet. Chem. 184
(1985) C33.
[33] P. Pons, C. Biran, M. Bordeau, J. Dunogues, J. Organomet. Chem. 358 (1988) 31.
[34] S. Bank, D.A. Juckett, J. Am. Chem. Soc. 98 (1976) 7742.
[35] G. Dabosi, M. Martineau, J. Simonet, J. Electroanal. Chem. 139 (1982) 211.
[36] V. Jouikov, C. Biran, M. Bordeau, J. Dunogues, Electrochim. Acta 45 (1999) 1015.
[37] V.V. Zhuikov, C. Biran, M. Bordeau, J. Dunogues, Rus. J. Gen. Chem. 67 (1997)
977.
[38] R.J. Corriu, Hypervalent Species of Silicon: Structure and Reactivity, in: A.R.
Bassindale, P.P. Gaspar (Eds.), Frontiers of Organosilicon Chemistry, RSC,
Cambridge, 1991.
[39] C. Biran, M. Bordeau, D. Bonafoux, D. Deffieux, D. Duprat, V. Jouikov, M.P. Leger-
Lambert, C. Moreau, F. Serein-Spirau, J. Chim. Phys. Phys. -Chim. Biol. 93 (1996)
591.
[54] C.P. Andrieux, E. Differding, M. Robert, J.M. Saveant, J. Am. Chem. Soc. 115
(1993) 6592.
[55] C.P. Andrieux, A. Le Gorande, J.-M. Saveant, J. Am. Chem. Soc 114 (1992) 6892.
[56] C.P. Andrieux, M. Robert, F.D. Saeva, J.-M. Saveant, J. Am. Chem. Soc. 116 (1994)
7864.
[57] A.A. Isse, C.Y. Lin, M.L. Coote, A. Gennaro, J. Phys. Chem. B 115 (2011) 678.
[58] J.-M. Saveant, J. Am. Chem. Soc. 109 (1987) 6788.
[59] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 105 (2005) 2999.
[60] L. Eberson, Acta Chem. Scand. Ser. B 36 (1982) 533.
[61] J.-M. Saveant, J. Am. Chem. Soc. 114 (1992) 10595.
[62] A. Fry, J. Electrochem. Commun. 7 (2005) 602.
[63] N.M. Emanuel, D.G. Knorre, Chemical Kinetics, 4th ed., Vysch. Shkola, Moscow,
1984.
[64] P.W. Atkins, Chimie Physique, De Boeck Universite s.a., Paris, 2000, pp. 835.
[65] M.G. Voronkov, V.M. Dyakov, S.V. Kirpichenko, J. Organomet. Chem. 233 (1982)
1.
[40] P. Clavel, G. Lessene, M. Bordeau, C. Biran, N. Roques, S. Trevin, D. De
Montauzon, J. Fluorine Chem. 107 (2001) 301.
[41] C. Moreau, F. Serein-Spirau, M. Bordeau, C. Biran, J. Dunogues, J. Organomet.
Chem. 522 (1996) 213.
[42] V. Jouikov, G. Salaheev, Electrochim. Acta 41 (1996) 2623.
[43] C.K. Mann, K.K. Barnes, Electrochemical Reactions in Nonaqueous Systems,
Marcel Dekker Inc, NY, 1970.
[44] M. Biran, Inorganic and Organometallic Polymers with Special Properties.
NATO ASI Ser. -Ser. E, in: R.M. Lain (Ed.), Kluwer Academic Publishers,
Dordrecht, The Netherlands, 1992.
[45] Gaussian 03, Revision B.01, M.J. Frisch, G.W., Trucks, H.B., Schlegel, G.E.,
Scuseria, M.A., Robb, J.R., Cheeseman, J.A., Montgomery, Jr., T. Vreven, K.N.,
Kudin, J.C., Burant, J.M., Millam, S.S., Iyengar, J., Tomasi, V., Barone, B.,
Mennucci, M., Cossi, G., Scalmani, N., Rega, G.A., Petersson, H., Nakatsuji, M.,
Hada, M., Ehara, K., Toyota, R., Fukuda, J., Hasegawa, M., Ishida, T., Nakajima, Y.,
Honda, O., Kitao, H., Nakai, M., Klene, X., Li, J.E., Knox, H.P., Hratchian, J.B., Cross,
C., Adamo, J., Jaramillo, R., Gomperts, R.E., Stratmann, O., Yazyev, A.J., Austin, R.,
Cammi, C., Pomelli, J.W., Ochterski, P.Y., Ayala, K., Morokuma, G.A., Voth, P.,
Salvador, J.J., Dannenberg, V.G., Zakrzewski, S., Dapprich, A.D., Daniels, M.C.,
Strain, O., Farkas, D.K., Malick, A.D., Rabuck, K., Raghavachari, J.B., Foresman, J.
V., Ortiz, Q., Cui A.G., Baboul, S., Clifford, J., Cioslowski, B.B., Stefanov, G., Liu A.,
Liashenko, P., Piskorz, I., Komaromi, R.L., Martin, D.J., Fox T., Keith, M.A. Al-
Laham, C.Y., Peng, A., Nanayakkara, M., Challacombe, P.M.W., Gill, B., Johnson,
[66] E. Herbert, J.P. Mazaleyrat, Z. Welvart, L. Nadjo, J.-M. Saveant, Nouv. J. Chem. 9
(1985) 75.
[67] O. Hammerich, V.D. Parker, Acta Chem. Scand. Ser. B 35 (1981) 341.
[68] D.R. Weyenberg, L.H. Toporcer, J. Am. Chem. Soc. 30 (1965) 943.
[69] K. Mai, G. Patil, J. Org. Chem. 51 (1986) 3545.
[70] D.R. Weyenberg, L.H. Toporcer, J. Am. Chem. Soc. 84 (1960) 2843.
[71] C. Hansch, A. Leo, R.W. Taft, Chem. Rev. 97 (1991) 165.
[72] C. Eaborn, R.A. Jackson, R. Pearce, J. Chem. Soc. Perkin Trans. I 17 (1974) 2055.
[73] L.H. Chan, E. Rochow, J. Organomet. Chem. 9 (1967) 231.
[74] E. Herbert, J.P. Mazaleyrat, Z. Welvart, J. Chem. Soc. Chem. Commun. 23 (1977)
877.
[75] G.N. Sastry, S. Shaik, J. Am. Chem. Soc. 117 (1995) 3290.
[76] J.Y. Corey, R. West, J. Am. Chem. Soc. 85 (1963) 4034.
[77] B. Suthar, A. Aldongarov, I.S. Irgibaeva, M. Moazzen, B.T. Donovan-Merkert, J.W.
Merkert, T.A. Schmedake, Polyhedron 31 (2012) 754.
[78] M. Dieng, D. Gningue-Sall, V. Jouikov, Main Group Met. Chem. 35 (2012) 141.
[79] C.M. Elliott, E.J. Hershenhart, J. Am. Chem. Soc. 104 (1982) 7519.
[80] J.K. Hino, L. Della Ciana, W.J. Dressick, B.P. Sullivan, Inorg. Chem. 31 (1992)
1072.
[81] H. Kunz, H. Waldman, in: B.M. Trost, I. Fleming (Eds.), Comprehensive Organic
Synthesis, 6, Pergamon, Oxford, 1991, pp. 631.