Tris(trimethylsilyl)silyl versus tris(trimethylsilyl)germyl: Radical reactivity and oxidation ability

Article abstract of DOI:10.1016/j.jorganchem.2008.08.039

A comparison of the tris(trimethylsilyl)silyl I and tris(trimethylsilyl)germyl II radical reactivity is provided. Their formation as well as their reactivity encountered in a large variety of chemical processes (addition to double bond, halogen abstraction, peroxyl radical formation...) is examined by laser flash photolysis, quantum mechanical calculations and electron spin resonance (ESR) experiments. The starting compound (TMS)₃GeH is more reactive than (TMS)₃SiH toward the t-butoxyl, the t-butylperoxyl and the phosphinoyl radicals. A similar behavior is noted for an aromatic ketone triplet state. II exhibits a lower absolute electronegativity: accordingly, the addition to electron rich alkenes is less efficient than for I. Radical II is also found less reactive for both the peroxylation (II + O₂ → II - O₂^{{radical dot}}) and the halogen abstraction reactions. The rearrangement of II - O₂^{{radical dot}} is slower than for I - O₂^{{radical dot}}; this is related to the respective exothermicity of the processes.

Full text of DOI:10.1016/j.jorganchem.2008.08.039

Journal of Organometallic Chemistry 693 (2008) 3643–3649
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Tris(trimethylsilyl)silyl versus tris(trimethylsilyl)germyl: Radical reactivity
and oxidation ability
b
a
a
Jacques Lalevée a, , Nicolas Blanchard , Bernadette Graff , Xavier Allonas ,
*
Jean Pierre Fouassier a
a Department of Photochemistry, UMR CNRS 7525, University of Haute Alsace, ENSCMu, 3 rue Alfred Werner, 68093 Mulhouse Cedex, France
b Department of Organic and Bioorganic Chemistry, UMR CNRS 7015, University of Haute Alsace, ENSCMu, 3 rue Alfred Werner, 68093 Mulhouse Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 July 2008
Received in revised form 28 August 2008
Accepted 29 August 2008
Available online 10 September 2008
A comparison of the tris(trimethylsilyl)silyl I and tris(trimethylsilyl)germyl II radical reactivity is pro-
vided. Their formation as well as their reactivity encountered in a large variety of chemical processes
(addition to double bond, halogen abstraction, peroxyl radical formation. . .) is examined by laser flash
photolysis, quantum mechanical calculations and electron spin resonance (ESR) experiments. The start-
ing compound (TMS)3GeH is more reactive than (TMS)3SiH toward the t-butoxyl, the t-butylperoxyl and
the phosphinoyl radicals. A similar behavior is noted for an aromatic ketone triplet state. II exhibits a
lower absolute electronegativity: accordingly, the addition to electron rich alkenes is less efficient than
for I. Radical II is also found less reactive for both the peroxylation ðII þ O2 ! II-OÅ2Þ and the halogen
abstraction reactions. The rearrangement of II-OÅ2 is slower than for I-OÅ2; this is related to the respective
exothermicity of the processes.
Keywords:
Radical chemistry
Silyl radical
Germyl radical
Laser flash photolysis
Oxidation
Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction
We have recently found that both structures can be used as
efficient co-initiators of free radical polymerization processes and
Since its introduction as an efficient radical-based reducing
agent more than twenty years ago, tris(trimethylsilyl)silane has
found multiple applications in organic synthesis as well as in poly-
mers and materials science [1–7]. Due to its low toxicity (com-
pared to n-tributyltin hydride) and the very mild reaction
conditions, tris(trimethylsilyl)silane has been a reagent of choice
for the reduction reaction of organic halides, selenides, activated
hydroxy- and carboxy groups even in aqueous media [1e]. The un-
ique reactivity of tris(trimethylsilyl)silane has also been exploited
in cascade reactions leading in a single step to complex molecular
scaffolds. Recent elegant examples for nature products synthesis
have been also reported [3]. In sharp contrast, tris(trimethyl-
silyl)germane has been much less exploited in organic synthesis
even though its reactivity is slightly superior to n-tributyltin hy-
dride [4]. Excellent yields were obtained in reduction reactions of
secondary or tertiary halides, isonitriles, nitro or seleno moieties.
Recently, tris(trimethylsilyl)germane was used in the radical-med-
iated germyldesulfonylation of vinylsulfones, thereby constituting
a straightforward access to vinylgermanes, reactive partners in
palladium-catalyzed cross-coupling reactions with aryl halides
[4b–c].
efficient photosensitizing species in free radical promoted cationic
photopolymerization. The intrinsic high reactivity of the generated
radicals associated with the rearrangement of the peroxyls formed
in aerated conditions (that regenerates the silyl or germyl radicals)
[5–7] allow highly efficient initiation processes even under air.
The aim of the present paper corresponds to an investigation of
the relative reactivity of tris(trimethylsilyl)silane (TMS)3Si–H and
tris(trimethylsilyl)germane (TMS)3Ge–H (Scheme 1); the tris(tri-
methylsilyl)methane (TMS)3C–H is partly introduced for compari-
son. The formation of the silyl and germyl radicals will be
investigated by laser flash photolysis and ESR spin trapping exper-
iments ESR-ST. The rate constants of interaction with different
additives (various monomers, lauraldehyde, diphenyliodonium
hexafluorophosphate, CBr4) will be determined to provide a coher-
ent picture of the SiÅ and GeÅ reactivity for a large range of pro-
cesses. Finally, the behavior under air of these compounds will
be investigated. Quantum mechanical calculations will help to
understand the observed reactivity.
2. Experimental
The starting compounds (tris(trimethylsilyl)silane (TMS)3SiH;
tris(trimethylsilyl)germane (TMS)3GeH; tris(trimethylsilyl)meth-
ane (TMS)3CH) were obtained from Aldrich and used with the best
purity available. In the case of liquid monomers – vinyl ethyl ether
* Corresponding author. Tel.: +33 3 89 33 68 37; fax: +33 3 89 33 68 95.
E-mail address: j.lalevee@uha.fr (J. Lalevée).
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.08.039

3644
J. Lalevée et al. / Journal of Organometallic Chemistry 693 (2008) 3643–3649
referred to the reversible formal potential of this compound
(+0.44 V/SCE).
2.5. Hydroperoxides formation under air
The formation of the hydroperoxides in benzophenone/
(TMS)3SiH or benzophenone/(TMS)3GeH in acetonitrile under light
irradiation (Hg–Xe lamp, 120 s, I ꢀ 20 mW/cm2) and under air was
studied through a classical iodometric determination [11]. After
irradiation, the concentration in hydroperoxides was determined
by addition of NaI through the following redox reaction sequence:
Scheme 1.
VE, vinyl acetate VA, methyl acrylate MA, acrylonitrile AN – (pur-
chased from Aldrich), the stabilizer (4-methoxyphenol) was re-
moved by column purification (Aldrich AL-154). Lauraldehyde
ROOH þ Iꢁ ! ROÅ þ OHꢁ þ I2
I2 þ Iꢁ ! Iꢁ3
(LA), diphenyliodonium hexafluorophosphate
(
U2I+), CBr4, 4-
The absorbance of the solutions was then measured with a UV–
Vis spectrometer (Beckman 640) at 358 nm (typical of Iꢁ3 ). In aceto-
nitrile, a molar extinction coefficient of 29400 Mꢁ1 cmꢁ1 was first
determined [11].
methoxyphenol (4-MeOP) and vitamin E (Vit E) were also obtained
from Aldrich.
2.1. Laser flash photolysis
2.6. DFT calculations
The silyl or germyl radical reactivity was studied by nanosecond
laser flash photolysis at RT (using the equipment described in [8];
resolution time: 10 ns). The solvent used is di-tert-butylperoxide/
benzene (50%/50%) and acetonitrile for the experiments with
tBu–OÅ and PÅ, respectively (see text below). The interaction of
(TMS)3GeH and (TMS)3SiH with a ketone triplet state (benzophe-
none BP) was studied in benzene; the ketyl radical quantum yield
UK. In the case of benzophenone was determined by a classical
procedure [8].
All the calculations were performed using the hybrid functional
B3LYP from GAUSSIAN 03 suite of program [12]. Reactants and prod-
ucts were fully optimized at the B3LYP/6-31+G level (and fre-
quency checked). For the calculation of the transition states (TS)
for the peroxyl radical rearrangement, the QST2 procedure, de-
scribed in detail in [12], was used. The peroxyl radicals absorption
*
*
spectra were calculated at TD(Nstates = 10)MPW1PW91/6-31+G
level.
2.2. Kinetic ESR (KSER)
3. Results and discussions
The ESR experiments were carried out using a X-band spec-
trometer (MS 200 from Magnettech-Berlin; Germany) at room
temperature. The radicals were generated through photolysis in
an air saturated inert solvent (tert-butylbenzene). During the pho-
tolysis, the spectrometer was set at the magnetic field correspond-
ing to the maximum peak height of the first derivative of the
observed radical and the field sweep was switched off. Decays in
the signal were monitored when the light was interrupted. The ki-
netic ESR (KESR) procedures have been described in detail in [9a–
e]. The interaction rate constants of tBu–OOÅ with (TMS)3GeH;
(TMS)3SiH and (TMS)3CH were determined from the lifetime of
tBu–OOÅ at different quencher concentrations through a classical
Stern–Volmer plot. The starting radical formed from the photolysis
of 2,2,4,4-tetramethylpent-3-one in an oxygenated medium is ob-
served at g = 2.015 [9a–b].
3.1. Generation of the silyl and germyl radicals
The generation of the silyl or germyl radicals can arise according
to different ways [13–16]. All these approaches are based on a
hydrogen abstraction reaction by the tert-butoxyl radical tBu–OÅ
(reaction 1 in Scheme 2) [13–14], the phosphinoyl radical PÅ (reac-
tion 2 in Scheme 2) or the benzophenone 3BP triplet state (reaction
3 in Scheme 2). A general review of the hydrogen donor abilities of
the group 14 hydrides has been presented in [15a]. The interaction
of tBu–OÅ with (TMS)3SiH and (TMS)3GeH as hydrogen donors D–H
yields I or II: their absorption (known as usually weak for
k > 300 nm) [15–16] can however be followed at 330 nm (Fig. 1).
The rise time of this absorption allows the evaluation of the inter-
action rate constants (Table 1). For reactions 2 and 3, the interac-
tion rate constants kH are extracted from the decay times of PÅ or
3BP observed at 450 and 525 nm in acetonitrile, respectively (Table
1 and Fig. 1). The phosphinoyl radical is easily generated by the di-
2.3. ESR spin trapping experiments
This ESR technique (ESR-ST) is now recognized as particularly
powerful for the identification of the radical centers [9f–g]. The
radicals generated under light exposure of a Xe–Hg lamp (Ham-
amatsu, L8252, 150 W; k > 310 nm) were trapped by phenyl-N-t-
butylnitrone (PBN). Tert-butylbenzene was used as a solvent.
ESR spectra simulations were carried out with the WINSIM software
[10].
2.4. Redox potentials
The redox potentials were measured in acetonitrile by cyclic
voltammetry with tetrabutyl-ammonium hexafluorophosphate
0.1 M as supporting electrolyte (Voltalab 06-Radiometer; the
working electrode was a platinium disk and the reference a satu-
rated calomel electrode-SCE). Ferrocene was used as a standard
and the potentials determined from the half peak potential were
Scheme 2.

J. Lalevée et al. / Journal of Organometallic Chemistry 693 (2008) 3643–3649
3645
higher reaction exothermicity for t-Bu–OÅ associated with the fact
that BDE(O–H) > BDE (P(O)–H) [18]. A very high reactivity of alkyl-
peroxyls with the Si–H and Ge–H functions is noted (Fig. 2 in Sup-
plementary material). The values found are different to those
obtained in [22] with a different approach (inhibited hydrocarbon
oxidation methodologies): 66 and 4000 Mꢁ1 sꢁ1 for (TMS)3SiH and
(TMS)3GeH, respectively. This method is also well known for the
evaluation of the ROOÅ interaction rate constants. This method is
based on a kinetic scheme. The reasons for the deviation between
KESR and the inhibited oxidation procedure are not known here.
Further works are probably necessary to understand this behavior.
The rate constants can also be advantageously compared to those
found for alkanes (k < 1 Mꢁ1 sꢁ1) [19]: this demonstrates the
intrinsic reactivity of both structures for the auto-oxidation pro-
cesses (see Section 3.4).
In the case of 3BP, the radical quantum yields are very high
(>0.8). The rate constant found with (TMS)3GeH (in the 109 Mꢁ1 sꢁ1
range) appears, however, as rather high for a pure hydrogen
abstraction reaction. This can be due to the lower oxidation poten-
tial of this specie (Eox = 1.44 V, Table 1) leading to an electron–pro-
ton transfer sequence for the overall hydrogen abstraction process
or at least to a high charge transfer character for the hydrogen
transfer. Indeed, the free energy change for a possible electron
transfer process
(DGet) can be evaluated from the classical
Rehm–Weller equation (Eq. (1), where Eox, Ered, ET and C are the
oxidation potential of the donor, the reduction potential of
the acceptor, the triplet state energy and the coulombic term for
the formed initial ion pair, respectively [20]; C was neglected here)
D
Get ¼ EOX ꢁ Ered ꢁ ET þ C
ð1Þ
Using for BP a reduction potential and a triplet energy level of
1.79 V and 2.98 eV, respectively [20b], the Get values of +0.25
D
and +0.51 eV are calculated for (TMS)3GeH and (TMS)3SiH, respec-
tively. From these results, it can be noted that, despite an ender-
Fig. 1. LFP experiments. (A) Decay traces monitored at 330 nm as a function of
[MA]; (a) for the germyl radical II ([MA] = 0 M) and (b) for the adduct II-MAÅ
([MA] = 0.062 M). (B) Decay trace of 3BP at [(TMS)3Ge-H] = 0.002 M.
gonic process, a quite low
DGet value is associated with the
germane derivative rendering feasible the electron transfer pro-
cess, in agreement with the high rate constant.
In comparison, the behavior of (TMS)3C–H was examined. The
hydrogen abstraction rate constants with tBuOÅ, tBuOOÅ, PÅ and
3BP are much lower than for Si–H and Ge–H (Table 1). This can
be ascribed both to a high BDE(C–H) (414.7 kJ/mol) and/or a higher
Eox rendering the hydrogen transfer or the electron proton transfer
sequence not exothermic enough. As the formation of this radical
cannot be observed, no other data are available.
The I and II radicals were also characterized by the ESR-ST tech-
nique for reactions 1–3 (Scheme 2). For these processes, the ad-
ducts were trapped with PBN, the corresponding hyperfine
splitting constants aH and aN are gathered in Table 1. It is particu-
larly worthnoting that aH is found much higher for II i.e. 9.85 ver-
sus 5.7 G for II and I, respectively.
rect cleavage of phenyl-bis(2,4,6-trimethylbenzoyl) phosphine
oxide at 355 nm [17]. For reaction 3, the ketyl radical quantum
yields UK (also equal to the quantum yields in I or II) are also given
in Table 1. The interaction rate constants of t-butylperoxyl tBuOOÅ
with (TMS)3SiH and (TMS)3GeH were obtained from KESR experi-
ments since alkylperoxyls absorb in a spectral window hardly
accessible by a LFP setup [13c].
Concerning the reaction with t-Bu–OÅ, t-BuOOÅ, PÅ and 3BP,
(TMS)3GeH is found more reactive than (TMS)3SiH with relative
rate constant ratios ꢀ5, 3, 50, 10, respectively. This can be ascribed
to the bond dissociation energy BDE (Ge–H) which is calculated
lower than the BDE (Si–H) thereby rendering the corresponding
abstraction process more exothermic (Table 1). A similar behavior
was noted for the reactivity of primary alkyl radicals with
(TMS)3SiH and (TMS)3GeH [15e]. The calculated BDE are in reason-
able agreement with the experimental data (Table 1) [18]. Interest-
ingly, PÅ is found more selective than t-Bu–OÅ. This is likely due to a
3.2. Structures of I and II
I and II have been characterized by DFT calculations. These rad-
icals exhibit a pyramidal structure as evidenced by the dihedral
Table 1
Rate constants and parameters characterizing the radical formation in benzene at RT
BDEa (kJ/mol)
kH (t-Bu–OÅ) (107 Mꢁ1 sꢁ1
)
kH (tBu–OOÅ) (Mꢁ1 sꢁ1
)
kH (PÅ) (107 Mꢁ1 sꢁ1
)
kH (3BP) (107 Mꢁ1 sꢁ1
)
UKÅ
Eox (V/SCE)
HFS (G)
I
II
III
334.0 (350.3)
316.8 (305.9)
414.7
8.5
45.0
<0.03
590
1790
<1
0.87
44
<0.1
10.2
105
<0.05
0.95
0.81
n.d.
1.7
1.44
>2
aN = 15.2; aH = 5.7
aN = 14.85; aH = 9.85
b
n.d. not determined (almost no quenching is observed).
a
Determined at UB3LYP/6-31+G* level; in brackets the experimental data from [18].
Not trapped.
b

3646
J. Lalevée et al. / Journal of Organometallic Chemistry 693 (2008) 3643–3649
angle (a) between the plan characterized by the two silicon atoms
R
S i
R
S i
R
and the radical center (Si or Ge) and the third silicon atom as de-
picted in Fig. 2. For a planar radical structure, must be 180°.
For I and II, is 156 and 150°, respectively. The bent out of the
R
+
R
a
R '
HC
R '
R
a
plane structure of the germyl radical is more pronounced. These re-
sults can be compared to the value found (171°) for (TMS)3CÅ. This
general property of the silyl and germyl radicals was already
known: the main difference between CÅ, SiÅ and GeÅ arises from
the fact that CÅ can only use 2s and 2p atomic orbitals to accommo-
date the valence electrons whereas SiÅ and GeÅ can use s, p and d
orbitals. The shapes of the silyl or germyl radicals are normally
Scheme 3.
determined from the decay time of I or II directly observed at
330 nm. The rate constants are gathered in Table 2.
In the addition of I and II to alkenes, a high reactivity is noted
(Table 2). For acrylonitrile AN and methyl acrylate MA (alkenes
bearing withdrawing substituents) a very similar reactivity is
found for I and II. For vinyl acetate VA and vinyl ether VE (electron
rich alkenes), I is more reactive than II. This can evidence the high-
er electrophilic character of I. This is in agreement with the higher
absolute electronegativity calculated for I (4.06 eV and 3.70 eV at
considered to be strongly bent out of the plane (r type struc-
ture).[1a] Interestingly, this behavior is enhanced for II. This can
be ascribed to the 3d orbitals which are already occupied and the
possible extension of the 4d orbital. For (TMS)3CÅ, an almost planar
structure is obtained. The deviation from planarity for CÅ can be due
to the steric hindrance associated with the TMS groups. The calcu-
lated spin density for the radical centers is 0.85; 0.98 and 1.1 for I,
II and III, respectively. This demonstrates that the maximal delo-
calization is observed for I.
*
UB3LYP/6-31+G level for I and II, respectively) [13–14].
As observed for the addition of other silyl radicals to a ketone
double bond [1a], it can be assumed that the addition to the C@O
bond of lauryl aldehyde LA occurs at the oxygen atom with the for-
mation of a carbon centered radical. A low reactivity is noted for I
and II (k < 6 ꢂ 106 Mꢁ1 sꢁ1); it can be compared to that found for
Ph3SiÅ and Ph3GeÅ: 6.2 ꢂ 107 and 9.5 ꢂ 106 Mꢁ1 sꢁ1, respectively.
The lower calculated addition exothermicity for I and II (ꢁ73.1
and ꢁ138.1 kJ/mol for I and Ph3SiÅ, respectively) is presumably
responsible of this behavior. For I, the addition to the oxygen
atom is found more exothermic than that for a carbon atom (by
53.4 kJ/mol) in agreement with the expected selectivity.
3.3. Reactivity of I and II toward different additives
The radicals were generated from the hydrogen abstraction
with t-BuOÅ. The reactivity of I and II was investigated by LFP for
different processes: (i) the addition to different double bonds
(C@C) or (C@O) (Scheme 3) (ii) the addition to oxygen leading to
a peroxyl radical formation, (iii) the halogen abstraction from
CBr4 and (iv) the oxidation of these radicals by U2IþðI þ U2Iþ
! Iþ þ UÅ
þ -IÞ. The comparison of the I/II reactivity for these dif-
U
In the halogen abstraction from CBr4, the rate constant is higher
for I (34 ꢂ 107 Mꢁ1 sꢁ1). This can be in line with the higher BDE
(Si–Br) compared to BDE (Ge–Br) rendering the abstraction process
more exothermic [18].
ferent processes can be useful to get a coherent picture of the
respective behavior of the silyl or germyl radicals.
The rate constants k for the processes (i) and (ii) were deter-
mined from the rising time of the adducts corresponding to these
reactions as the adducts absorb much more strongly than the start-
ing radical (Fig. 1) [14a]. For the processes (iii) and (iv), they were
In the radical oxidation with U2I+, a quite high rate constant
(0.26 ꢂ 107 Mꢁ1 sꢁ1) is obtained with I in agreement with the
low ionization potential IP usually assumed for silyl radicals [1a].
For II, an upper value of 2 ꢂ 107 Mꢁ1 sꢁ1 can only be given: the
degradation of the solution under light irradiation prevents a more
refined analysis. The IP of II is calculated lower than for I (6.0 ver-
sus 6.36 eV for I): this likely demonstrates that germyl radicals can
also be oxidized by an aryliodonium salt. This result is in agree-
ment with the photosensitization of a cationic polymerization re-
cently observed when using these radicals [7].
3.4. Oxidation process of (TMS)3SiH and (TMS)3GeH
3.4.1. Laser flash photolysis investigation
Under air, the tBuOÅ/(TMS)3SiH and tBuOÅ/(TMS)3GeH interac-
tions lead to new transients with a maximum absorption at 420
and 390 nm, respectively (Figs. 3 and 4). From the measurement
of the radical rising time at various [O2], the interaction rate con-
Table 2
Rate constants k for the reaction of the radicals I and II with various additives in di-
tert-butylperoxide
I
II
107 k (Mꢁ1 sꢁ1
)
107 k (Mꢁ1 sꢁ1
)
AN
MA
VA
VE
LA
5.1
2.2
4.5
3.4
<0.05
<0.05
<0.6
39
0.12
0.02
<0.6
200
34
O2
CBr4
8
<2
*
Fig. 2. Structures of the radicals: (A) I, (B) II, (C) (TMS)3CÅ at UB3LYP/6-31+G level.
U2I+
0.26
The definition used for the dihedral angle
a is also given.

J. Lalevée et al. / Journal of Organometallic Chemistry 693 (2008) 3643–3649
3647
Interestingly, the lifetime of I-OÅ2 is found short (about 2.5
ls)
compared to II-OÅ2 (>5 ms). For I-OÅ2, it has been already suspected
that a fast rearrangement (competitive to the possible formation
of a hydroperoxide) occurs with the migration of a trimethylsilyl
(TMS) substituent (Scheme 4 as proposed in [22] albeit no direct ki-
netic information is available) leading to a silyloxyl radical A and
then a silyl radical B. This Scheme was originally proposed in agree-
ment with 16O2, 18O2 labeling experiments [22]. This rearrangement
appears less favorable for the germyl structure as II-OÅ2 has a much
higher lifetime (Fig. 3 in Supplementary material).
To shed some light on this behavior, the associated reorganiza-
*
tion enthalpies have been calculated at UB3LYP/6-31+G level.
Interestingly, it is clearly found that the TMS group migration pro-
cess is more exothermic for I-OÅ2 (ꢁ462 kJ/mol) compared to II-O2Å
(ꢁ340 kJ/mol) in agreement with the respective peroxyl stability.
The basic difference is the lower BDE (Ge–O) compared to BDE
(Si–O) calculated as 397.6 and 446.1 kJ/mol for CH3Ge–OCH3 and
*
CH3Si–OCH3 (at UB3LYP/6-31+G level). The transition states cor-
responding to the TMS group migration (reaction from the metal
peroxyl M–O–OÅ to the structure noted A in Scheme 4) were calcu-
lated; the TS structures are given in the Supplementary material.
Å
*
The barrier calculated at UB3LYP/6-31+G is found lower for I-O2
(50.1 kJ/mol) than for II-OÅ2 (122.2 kJ/mol). This is in full agreement
with the shorter lifetime of I-OÅ2.
The direct observation of the peroxyl radicals can be valuable
for a direct access to their chemical reactivity. Due to its short life-
time, I-OÅ2 is not a convenient probe for the evaluation of the low
rate constants (upper values in the 107 Mꢁ1 sꢁ1 range can only be
obtained). The interaction with different additives (4-methoxy
phenol, vitamin E and U2I+) has been investigated using II-O2Å
which is characterized by a much longer lifetime (Table 3). The
reaction with 4-MeOP or vitamin E is quite efficient (about 10
times higher than that noted for alkyl peroxyls) [13c,21]. The same
behavior is found for U2I+. This evidences the high reactivity of this
metal peroxyl radical.
Fig. 3. Transient peroxyl radical spectra associated with I (A) and II (B) in di-tert-
butylperoxide (oxygen saturated solution, [(TMS)3SiH] = 0.05 M;
[(TMS)3GeH] = 0.05 M). The calculated peroxyl radical spectra are given (black
*
bar): kmax and oscillator strength (f) at TD-DFT/MPW1PW91/6-31+G level.
3.4.2. ESR spectra under air
The ESR results obtained under air confirm the information
gained through LFP. For I, a spectrum characterized by g ꢃ 2.0045
is recorded (Fig. 5) and is safely assigned to the final silyl radical
(specie B in Scheme 4) the g factor being in the typical range
known for SiÅ radicals [1a]. Moreover, this g value is too low for a
peroxyl radical (usually in the 2.011–2.015 range for C–OOÅ and
2.020–2.025 for Si–OOÅ) [23]. For II, a specie characterized by
Scheme 4.
Fig. 4. Building up of the transient II-OÅ2 absorption in di-tert-butylperoxide (a)
under argon, (b) in aerated solution ([O2] = 0.0019 M) and (c) in saturated oxygen
solution ([O2] = 0.0091 M).
Table 3
Rate constants k for the reaction of the peroxyl radicals (derived from II) with various
additives in di-tert-butylperoxide
stants for I/O2 and II/O2 are determined as 2 ꢂ 109 and
3.9 ꢂ 108 Mꢁ1 sꢁ1. The observed new transients are ascribed to
the corresponding peroxyl radicals as the calculated and experi-
mental absorption spectra are in good agreement (Fig. 3). It is
worthnoting that the absorption of these peroxyls are red shifted
compared to the classical absorption of alkylperoxyl radicals
(k < 270 nm) [13c,13d,21]: this is ascribed to a partial charge trans-
fer transition.
II-O2Å 106 k (Mꢁ1 sꢁ1
)
Rearrangement
<0.001a (0.00001b)
k(4MeOP)
0.11
5.5
k(Vit
E)
k(
U
1.0
2I+)
a
In sꢁ1
KESR data.
.
b

3648
J. Lalevée et al. / Journal of Organometallic Chemistry 693 (2008) 3643–3649
4. Conclusion
In the present paper, the reactivity of two selected silane and
germane compounds was examined and compared through the
determination of the reaction rate constants of the processes.
Among the gathered new information, the oxidation behavior
was specifically investigated: it underlines the rearrangement of
the peroxyls which is concomitant with a low hydroperoxide for-
mation. All this new approach based on a direct access to the rad-
icals reactivity (as the Si and Ge peroxyl radicals were for the first
time observed by LFP) should be useful for a better understanding
of the associated behavior under air. The use of silyl or germyl rad-
icals as polymerization initiating species in aerated media will be
discussed in forthcoming papers.
Acknowledgments
The authors thank the CINES (Centre Informatique National de
l’Enseignement Supérieur) and IDRIS (Institut du Développement
et des Ressources en Informatique Scientifique-CNRS) for the gen-
erous allocation of time on the IBM SPsupercomputer.
Appendix A. Supplementary material
(i) Use of (TMS)3SiH as reported by Nicolaou in his synthetic ef-
forts towards azadirachtin [3]. (ii) Transition states corresponding
to the TMS group migration (reaction from the metal peroxyl
M–O–OÅ to the structure noted A in Scheme 4). (iii) KESR experi-
ments for the t-BuO–OÅ/(TMS)3SiH interaction. (iv) Peroxyl radical
decays in LFP experiments: I-OOÅ and II-OOÅ. Supplementary data
associated with this article can be found, in the online version, at
doi:10.1016/j.jorganchem.2008.08.039.
Fig. 5. ESR spectra recorded under air in di-tert-butylperoxide for (a) (TMS)3GeH
(0.02 M); (b) (TMS)3SiH (0.02 M); (c) (TMS)3CH (0.040 M). The g calibration was
carried out with tetramethylpiperidine N-Oxyl (TEMPO) as a reference. Insert:
decay of II-OÅ2 at 3306.5 G observed by kinetic ESR (KESR).
g = 2.021 is observed: this high g value is typical of a germylperoxyl
structure (i.e. for the trimethylgermylperoxyl, a broad singlet ESR
spectrum centered at g = 2.0245 was already observed in [23]).
The decay of II-OÅ2 observed by KESR (Fig. 5) can be fitted by an
exponential decay (radical lifetime: 99 ms). This leads to a reorga-
nization rate constant of ꢀ10 sꢁ1 (Scheme 4). This is in agreement
with the LFP results for which a much shorter lifetime for I-OÅ2 was
observed. Due to the time resolution of our equipment (actually
about 1 ms), the decay of I-OÅ2 cannot be followed here.
References
[1] (a) C. Chatgilialoglu, in: Organosilanes in Radical Chemistry, Wiley, Chichester,
2004;
(b) J. Fossey, D. Lefort, J. Sorba, in: Free Radicals in Organic Chemistry,
Wiley&Sons, New York, 1995;
(c) S.Z. Zard, in: Radical reactions in Organic Synthesis, Oxford University
Press, Oxford, UK, 2003;
For the tBuOÅ/(TMS)3CH system under air, the specie detected is
typical of a carbon centered radical (Fig. 5) [23]. This result evi-
dences the formation of (TMS)3CÅ. As the formation of peroxyls is
not observed, a low reactivity of (TMS)3CÅ toward O2 is expected.
This is in contrast with the almost diffusion controlled reactions
(ꢀ3–4 ꢂ 109 Mꢁ1 sꢁ1) often observed for the addition of carbon
centered radicals to O2. This behavior is probably ascribed to the
steric hindrance associated with a CÅ radical bearing TMS bulky
substituents.
(d) C. Chatgilialoglu, in: P. Renaud, M.P. Sibi (Eds.), Radicals in Organic
Synthesis, vol.1, Wiley-VCH, Weinheim, Germany, 2001, pp. 28–49;
(e) A. Postigo, S. Kopsov, C. Ferreri, C. Chatgilialoglu, Org. Lett. 9 (2007) 5159.
[2] (a) C. Chatgilialoglu, Chem. Eur. J. 14 (2008) 2310;
(b) C. Chatgilialoglu, Acc. Chem. Res. 25 (1992) 188;
(c) C. Chatgilialoglu, A. Guerrini, M. Lucarini, J. Org. Chem. 57 (1992) 3405;
(d) M. Ballestri, C. Chatgilialoglu, K.B. Clark, D. Griller, B. Giese, B. Kopping, J.
Org. Chem. 56 (1991) 678.
[3] K.C. Nicolaou, P.K. Sasmal, T.V. Koftis, A. Converso, E. Loizidou, F. Kaiser, A.J.
Roecker, C.C. Dellios, X.-W. Sun, G. Petrovic, Angew. Chem., Int. Ed. 44 (2005)
3447.
[4] (a) C. Chatgilialoglu, M. Ballestri, Organometallics 14 (11) (1995) 5017;
(b) S.F. Wnuk, P.I. Garcia Jr., Z. Wang, Org. Lett. 6 (2004) 2047;
(c) Z. Wang, S.F. Wnuk, J. Org. Chem. 70 (2005) 3281.
[5] J. Lalevée, M. El-Roz, F. Morlet-Savary, B. Graff, X. Allonas, J.P. Fouassier,
Macromolecules 40 (2007) 8527.
[6] J. Lalevée, A. Dirani, M. El-Roz, X. Allonas, J.P. Fouassier, Macromolecules 41
(2008) 2003.
[7] J. Lalevée, A. Dirani, M. El-Roz, X. Allonas, J.P. Fouassier, J. Polym. Sci. Part A:
Polym. Chem. 46 (2008) 2008.
3.4.3. Hydroperoxide formation
The formation of hydroperoxides in BP/(TMS)3SiH,
BP/(TMS)3GeH, BP/Ph3SiH and BP/Ph3GeH in acetonitrile upon light
irradiation under air was investigated. Contrary to BP/Ph3SiH and
BP/Ph3GeH, the Iꢁ anion is not detected in BP/(TMS)3SiH and
3
BP/(TMS)3GeH: the hydroperoxide formation ðI-OÅ2 þ I-H ! IÅþ
I-O2HÞ is low (<1% to the quantity observed for Ph3SiH or Ph3GeH)
and can be neglected. This is a strong evidence for the competitive
rearrangement of I-OÅ2 and II-OÅ2. This also demonstrates that the
reorganization process for II-OÅ2 is the major pathway governing
the decay of this specie despite a quite low rate constant. In the
case of Ph3SiH and Ph3GeH, a phenyl migration in the correspond-
ing peroxyls is probably not favourable and hydroperoxides are ob-
served. This is in agreement with the higher lifetime for Ph3GeOOÅ
observed here by KESR at g = 2.028 (>2 s and decaying by a second
order kinetics supporting the usual recombination process).
[8] J. Lalevée, X. Allonas, J.P. Fouassier, J. Am. Chem. Soc. 124 (2002) 9613.
[9] (a) E. Furimsky, J.A. Howard, J. Am. Chem. Soc. 95 (1973) 369;
(b) B.W. Burton, T. Doba, E.J. Gabe, L. Hughes, F.L. Lee, L. Prasad, K.U. Ingold, J.
Am. Chem. Soc. 107 (1985) 7053;
(c) S. Fukuzumi, K. Shimoosako, T. Suenobu, Y. Watanabe, J. Am. Chem. Soc.
125 (2003) 9074;
(d) J.A. Howard, Landolt Bornstein Numerical Data and Functional
Relationships in Science and Technology, Group II, vol. 18, sub vol. D2,
Springer-Verlag, Berlin, 1997;
(e) D. Griller, J.A. howard, P.R. Marriott, J.C. Scaiano, J. Am. Chem. Soc. 103
(1981) 619;
(f) P. Tordo, in: N.M. Atherton, M.J. Davies, B.C. Gilbert (Eds.), Electron Spin
Resonance, vol. 16, The Royal Society of Chemistry, Cambridge, 1998;
(g) Y. Kotake, K. Kuwata, Bull. Chem. Soc. Jpn. 54 (1981) 394.

J. Lalevée et al. / Journal of Organometallic Chemistry 693 (2008) 3643–3649
3649
[10] D.R. Duling, J. Magn. Reson. Ser. B 104 (1994) 105.
[11] D.S. Breslow, D.A. Simpson, B.D. Kramer, R.J. Schwarz, N.R. Newburg, Ind. Eng.
Chem. Res. 26 (1987) 2144.
(d) C. Chatgilialoglu, S. Rossini, Bull. Soc. Chim. Fr. 2 (1988) 298;
(e) C. Chatgilialoglu, M. Ballestri, J. Escudie, I. Pailhous, Organometallics 18
(1999) 2395.
[12] (a) GAUSSIAN 03, Revision B-2, M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E.
Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., R.E.
Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K.
Morokuma, P. Salvador, J.J. Dannenberg, D.K. Malick, A.D. Rabuck, K.
Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, 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, W. Chen, M.W. Wong, J.L. Andres, C.
Gonzalez, M. Head-Gordon, E.S. Replogle, J.A. Pople, GAUSSIAN, Inc., Pittsburgh
PA, 2003;
[16] (a) C. Chatgilialoglu, M. Guerra, A. Guerrini, G. Seconi, K.B. Clark, D. Griller, J.
Kanabus-Kaminska, J.A. Martinho-Simoes, J. Org. Chem. 57 (1992) 2427;
(b) C. Chatgilialoglu, K.U. Ingold, J. Lusztyk, A.S. Nazran, J.C. Scaiano,
Organometallics 2 (1983) 1332.
[17] K. Dietliker, A Compilation of Photoinitiators Commercially Available for UV
Today, Sita Technology Ltd., Edinburgh, London, 2002.
[18] (a) Y.R. Luo, Handbook of Bond Dissociation Energy in Organic Compounds,
CRC Press, 2003;
(b) R. Beccera, R. Walsh, in: Z. Rappoport, Y. Apeloig (Eds.), The Chemistry of
Organic Silicon Compounds, vol. 2, John Wiley & Sons, 1994 (Chapter 4);
(c) T.I. Drozdova, E.T. Denisov, Kinet. Catal. 43 (2002) 14.
[19] J.A. Howard, J.C. Scaiano, in: H. Fischer (Ed.), Landolt–Bornstein new series,
group II, vol. 13, Radical Reaction Rates in Liquids, Part D, Springer-Verlag,
Berlin, 1984, pp. 251–253.
(b) J.B. Foresman, A. Frisch, in: Exploring Chemistry with Electronic Structure
Methods, 2nd ed., GAUSSIAN. Inc., 1996.
[13] (a) J. Lalevée, X. Allonas, B. Graff, J.P. Fouassier, J. Phys. Chem. A 111 (2007)
6991;
[20] (a) D. Rehm, A. Weller, Israel J. Chem. 8 (1970) 259;
(b) S.L. Murov, I. Carmichael, G.L. Hug, Handbook of Photochemistry, Ed.
Marcel Dekker Inc., New York, 1993.
(b) J. Lalevée, X. Allonas, J.P. Fouassier, Chem. Phys. Lett. 415 (2005) 202;
(c) J. Lalevée, X. Allonas, J.P. Fouassier, Chem. Phys. Lett. 445 (2007) 62;
(d) J. Lalevée, X. Allonas, J.P. Fouassier, K.U. Ingold, J. Org. chem. 73 (2008)
6489.
[21] K.U. Ingold, Acc. Chem. Res. 2 (1969) 1.
[22] (a) A.B. Zaborovskiy, D.S. Lutsyk, R.E. Prystansky, V.I. Kopylets, V.I. Timokhin,
C. Chatgilialoglu, J. Organomet. Chem. 689 (2004) 2912;
(b) C. Chatgilialoglu, A. Guarini, A. Guerrini, C.G. Seconi, J. Org. Chem. 57
(1992) 2207;
(c) C. Chatgilialoglu, A. Guerrini, M. Lucarini, G.F. Pedulli, P. Carrozza, G. Da
Roit, V. Borzatta, V. Lucchini, Organometallics 17 (1998) 2169.
[23] (a) J.A. Howard, J.C. Tait, S.B. Tong, Can. J. Chem. 57 (1979) 2761;
(b) J.A. Howard, J.C. Tait, Can. J. Chem. 54 (1976) 2669;
(c) J.A. Howard, J.C. Tait, Can. J. Chem. 56 (1978) 2163.
[14] (a) J. Lalevée, X. Allonas, J.P. Fouassier, J. Org. Chem. 72 (2007) 6434;
(b) J. Lalevée, X. Allonas, F. Morlet-Savary, J.P. Fouassier, J. Phys. Chem. A 110
(2006) 11605.
[15] (a) C. Chatgilialoglu, M. Newcomb, Adv. Organomet. Chem. 44 (1999) 67;
(b) C. Chatgilialoglu, K.U. Ingold, J.C. Scaiano, J. Am. Chem. Soc. 105 (1983)
3292;
(c) C. Chatgilialoglu, K.U. Ingold, J.C. Scaiano, J. Am. Chem. Soc. 104 (1982)
5119;