Silyloxyamines as sources of silyl radicals: ESR spin-trapping, laser flash photolysis investigation, and photopolymerization ability

Article abstract of DOI:10.1002/poc.1762

Two silyloxyamines derived from 8-(pentamethyldisilyloxy)-julolidine and diethyl 3-(pentamethyldisilyloxy)-aniline are proposed as new sources of silyl radicals. The decomposition mechanism, excited state processes and the radical generation are explored

Full text of DOI:10.1002/poc.1762

Research Article
Received: 5 February 2010,
Revised: 27 May 2010,
Accepted: 1 June 2010,
Published online 23 July 2010 in Wiley Online Library: 2011
(wileyonlinelibrary.com) DOI 10.1002/poc.1762
Silyloxyamines as sources of silyl radicals:
ESR spin-trapping, laser flash photolysis
investigation, and photopolymerization
ability
a
a
a
*
´
Davy Louis Versace , Mohamad Ali Tehfe , Jacques Lalevee ,
Virginie Casarottob, Nicolas Blanchardb, Fabrice Morlet-Savarya
and Jean-Pierre Fouassiera
Two silyloxyamines derived from 8-(pentamethyldisilyloxy)-julolidine and diethyl 3-(pentamethyldisilyloxy)-aniline
are proposed as new sources of silyl radicals. The decomposition mechanism, excited state processes and the radical
generation are explored by steady state photolysis, laser flash photolysis (LFP), electron spin resonance (ESR), and MO
calculations. The Si—Si bond cleavage is clearly demonstrated. The formation of a radical cation on the amine moiety
is also observed. Moreover, these compounds work as efficient Type I and Type II photoinitiators (PI) of free radical
photopolymerization (FRP). Copyright ß 2010 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: ESR spin-trapping; laser flash photolysis; photoinitiators; photopolymerization; silyl radicals
great interest. The aim of the present paper is to introduce new
sources of silyl radicals based on the design of silyloxyamines
INTRODUCTION
The development of silyl functionalized compounds and silyl
radical sources is now the subject of many efforts in organic
chemistry.[1,2–4] In particular, the bulky tris(trimethylsilyl)silyl
group has been found to promote [2þ2] cyclizations[3] and
Mukaiyama aldol reactions.[4] In addition, the silyloxy compounds
derived from tris(trimethylsilyl)silane (‘super silyl ethers’) can
be used as interesting UV-labile silyl protecting groups of
alcohols[1,2–4] (Scheme 1 where R stands for a large diversity of
substituents). This photocleavage provides an alternative
approach to the classical thermal deprotection.[5] However, little
is known on the photochemical mechanisms associated with
such a reaction; this process can involve photogeneration of silyl
radicals.
having easy and convenient synthetic procedures. The new
proposed structures could be useful for FRP. The limited access
and stability of phenoxy tris(trimethylsilyl)silanes led us to
investigate alternative compounds such as new silyloxyamines
derived from pentamethyldisilane.[5] Two starting structures
(8-hydroxyjulolidine and 3-diethylaminophenol) have been
selected for the synthesis of the corresponding silyloxyamines
(Scheme 2): silyloxyjulolidine (SiN1), silyloxyaniline (SiN2). In
contrast with other silyl sources previously developed and built
on a carbonyl chromophore, these molecules are expected to
possess a charge transfer transition (the donor being the amine
and the acceptor the silyl ether group) allowing a tunable
red-shift of the absorption through a relatively easy introduction
of various amine skeletons. This could help in the design of
efficient longer wavelength absorbing PIs for less toxic irradiation
lamps.[15,16]
The silyl radical chemistry which is now the subject of
numerous studies[6–16] appears highly worthwhile for the
development of new mono- and bi-component sources of silyl
.
radicals R3Si usable as photoinitiators (PI) for free radical
photopolymerization (FRP): indeed, silyl radicals are highly
valuable to overcome the oxygen inhibition usually observed
.
in FRP.[9–16] In this area, the production of R3Si was first achieved
´
Correspondence to: J. Lalevee, Department of Photochemistry, CNRS, Univer-
sity of Haute Alsace, ENSCMu, 3 rue Alfred Werner, 68093 Mulhouse Cedex,
France.
E-mail: j.lalevee@uha.fr
*
through the photodegradation of polysilane polymers that
occurs via a Si—Si bond cleavage[6–8] or the use of photo-
sensitizer/silane couples. New compounds based on Si—Si, C—Si,
.
or S—Si single bond cleavage and leading to R3Si have also been
´
a
D. L. Versace, M. A. Tehfe, J. Lalevee, F. Morlet-Savary, J.-P. Fouassier
recently proposed.[9–16] On this occasion, their ability as Type I
photoinitiating systems to initiate both FRP and free radical
promoted cationic polymerization (FRPCP) was underlined.
The design and the development of new PIs from silyl
containing compounds exhibiting novel properties remain of
Department of Photochemistry, CNRS, University of Haute Alsace, ENSCMu,
3 rue Alfred Werner, 68093 Mulhouse Cedex, France
b
V. Casarotto, N. Blanchard
Laboratory of Organic and Bioorganic Chemistry, CNRS, University of Haute
Alsace, ENSCMu, 3 rue Alfred Werner, 68093 Mulhouse Cedex, France
J. Phys. Org. Chem. 2011, 24 342–350
Copyright ß 2010 John Wiley & Sons, Ltd.

SILYLOXYAMINES AS SOURCES OF SILYL RADICALS
material. The estimated purity is >95%. From the UV spectra and
taking into account that the visible absorption of the starting
compound (8-hydroxyjulolidine) completely disappears; the pre-
sence of this starting compound in the final purified product is
<0.5%.
1H NMR (CDCl3, 400 MHz, 298 K) d 6.64 (d, J ¼ 7.9 Hz, 1H); 6.07
(d, J ¼ 7.9 Hz, 1H); 3.08 (p, J ¼ 5.5 Hz, 4H), 2.71 (t, J ¼ 6.5 Hz, 2H);
2.63 (t, J ¼ 6.5 Hz, 2H); 1.99–1.92 (4H); 0.36 (s, 6H); 0.10 (s, 9H). 13C
NMR (100 MHz, CDCl3, 298 K) d 152.2; 146.1; 126.2; 114.7; 112.5;
106.6; 50.2; 49.8; 27.2; 22.4; 21.94; 21.86; 2.08; ꢀ2.2.
Si
hv (254 nm)
R O Si Si
Si
R OH
Scheme 1.
Through laser flash photolysis (LFP), electron spin resonance
spin-trapping (ESR-ST), and quantum mechanical calculations,
we will evaluate the silyl radical formation upon UV irradiation,
investigate the photopolymerization ability of these silyloxya-
mines by real-time FTIR, and propose a mechanism for the
initiation step of the acrylate photopolymerization.
Synthesis of SiN2
To a solution of 3-diethylaminophenol (0.15 g, 0.9 mmol) in
CH2Cl2 (1 ml) were added DMAP (10 mol%), i-PrNEt2 (0.32 ml,
1.8 mmol), and chloropentamethyldisilane (0.2 ml, 1 mmol).
Monitoring the progress of the reaction by silica gel, thin layer
chromatography was difficult as SiN2 is prone to desilylation on
prolonged contact with SiO2. Therefore, after 12 h the reaction
mixture was concentrated under vacuum and the residual oil was
adsorbed on neutral alumina using CH2Cl2. The dried solid was
then loaded on a small pad of silica gel and eluted quickly with
cyclohexane/ethyl acetate ¼ 100:5. The desired product eluted
first (clearly different from 3-diethylaminophenol by alumina
coated thin layer chromatography analysis) and was obtained as
colorless oil (0.21 g, 79%). A copy of the 1H and 13C NMR spectra is
provided as supplementary material. The estimated purity is
>95%. From the UV spectra and taking into account that the
visible absorption of the starting compound (3-diethylamino-
phenol) completely disappears; the presence of this starting
compound in the final purified product is <0.5%.
EXPERIMENTAL PART AND
COMPUTATIONAL PROCEDURE
Samples
8-hydroxyjulolidine and 3-diethylaminophenol were purchased
from Aldrich and used with the best purity available. Chloro-
pentamethyldisilane was purchased from Aldrich and used as
received. Benzophenone (BP) was used as reference Type II
photoinitiator (Aldrich). Ethyldimethylaminobenzoate (EDB-Esacure
from Lamberti) and Tris(trimethylsilyl)silane (TTMSS – Aldrich)
were chosen as reference co-initiators.
Synthesis procedures for SiN1 and SiN2
1H NMR (CDCl3, 300 MHz, 298 K) d 7.03 (t, J ¼ 8.1 Hz, 1H); 6.30
(d, J ¼ 7.2 Hz, 1H); 6.17–6.12 (2H); 3.31 (q, J ¼ 6.9 Hz, 4H); 1.15
(t, J ¼ 6.9 Hz, 6H); 0.37 (s, 6H), 0.08 (s, 9H). 13C NMR (75 MHz, CDCl3,
298 K) d 157.2; 149.2; 129.6; 107.0; 105.4; 103.7; 44.3; 12.6;
0.2; ꢀ2.2.
Synthesis of SiN1
To a solution of 8-hydroxyjulolidine (0.1 g, 0.53 mmol) in CH2Cl2
(0.5 ml) were added 4-dimethylaminopyridine (DMAP) (10 mol%,
6.5 mg), i-PrNEt2 (0.18 ml, 1.06 mmol), and chloropentamethyldi-
silane (0.16 ml, 0.84 mmol). After 48 h, a conversion of 70% was
reached as indicated by thin layer chromatography (TLC) analysis
(cyclohexane/ethyl acetate ¼ 85:15). The reaction mixture was
concentrated under vacuum and the residual oil was adsorbed on
silica gel using CH2Cl2. The dried solid was then loaded on silica
gel and eluted with cyclohexane/ethyl acetate ¼ 100:3 following
the rapid chromatographic technique (see supporting infor-
mation). The desired product eluted first (clearly different from
8-hydroxyjulolidine by silica coated thin layer chromatography
analysis) and was obtained as colorless oil (0.10 g, 59%). A copy
of the 1H and 13C NMR spectra is provided as supplementary
Polymerization experiments
For the film photopolymerization (20 mm thick film), a given
silyloxyamine photoinitiator (3% w/w; weight concentration with
respect to the monomer) was dissolved in the polymerizable
triacrylate monomer (trimethylolpropane triacrylate TMPTA from
Cytec). The laminated or oxygenated films deposited on a BaF2
chip were irradiated with the polychromatic light of an Xe
lamp (filtered at l > 390 nm; light intensity ¼ 60 mW/cm2 in the
390 < l < 800 nm range) or a polychromatic light of a Xe–Hg
lamp (L8252, hamamatsu; 150 W; light intensity ¼ 44 mW/cm2 in
the 300–400 nm range). The polymerization kinetics of TMPTA
were examined by measuring the consumption of monomer in
photoinitiated experiments. The polymerization efficiencies of
the photoinitiating systems were studied by measuring the
percent conversion of TMPTA as a function of time. The evolution
of the monomer double bond absorption at about 1630 cmꢀ1
Si
Si
Si
Si
O
O
was continuously followed by real-time FTIR spectroscopy (Nexus
[13–17]
870, Nicolet) as described in References
therein.
and references
N
N
Silyloxyamines absorption spectra
The UV-visible absorption spectra (200–600 nm) in acetonitrile
were performed in a Beckman DU-640 Spectrophotometer.
Solutions were filled into a 1-cm UV path cell.
SiN2
SiN1
Scheme 2.
J. Phys. Org. Chem. 2011, 24 342–350
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

D. L. VERSACE ET AL.
Laser flash photolysis (LFP)
The nanosecond transient absorption setup working at 355 nm is
based on a nanosecond Nd:YAG laser (Powerlite 9010, Con-
tinuum) operating at 10 Hz with an energy decreased down to
about 10 mJ/pulse at 266 nm. The analyzing system (LP900,
Edinburgh Instruments) used a 450-W pulsed xenon arc lamp,
a Czerny–Turner monochromator, a fast photomultiplier, and a
transient digitizer (TDS 340, Tektronix). The beam irradiated a
1-cm cell containing the PI. Measurements were done at room
temperature. Before each measurement, solutions were
degassed by argon bubbling.[10–12]
ESR spin-trapping
ESR-ST experiments were carried out using an X-Band spec-
trometer (MS 200 Magnettech). The radicals generated under
light irradiation (Xe–Hg lamp – Hamamatsu L8252, 150 W;
l > 310 nm) were trapped by phenyl-N-tbutylnitrone (PBN, Assay
ꢁ98%, Aldrich). This technique is useful for the identification of
the radical center.[18–20] The ESR spectra simulations were carried
out with the PEST WINSIM program.[21]
Cyclic voltammetry
The redox potentials were measured in acetonitrile by cyclic
voltammetry with tetrabutyl ammonium hexafluorophosphate
(0.1 M) as a supporting electrolyte. The working electrode was
a platinum disc and the reference one, a saturated calomel
electrode (SCE). Ferrocene was used as a standard reference
(Eox ¼ þ 0.44 V/SCE). The free-energy changes (DGEt) for the
electron transfer from the free radical to the triplet energy
acceptor can be calculated from the classical Rehm–Weller
equation[22] (Eqn 1) where Eox is the oxidation potential of
the donor, Ered the reduction potential of the acceptor, ET
the triplet state energy, C a coulombic term for the initially
formed ion pair (C is neglected here as usually done in polar
solvents).
Figure 1. UV absorption properties: (A) 8-hydroxyjulolidine (dash line)
and SiN1 (solid line); inset: zoom between 300 and 600 nm.
(B) 3-diethylaminophenol (dash line) and SiN2 (solid line); in acetonitrile
DGEt ¼ Eox ꢀ Ered ꢀ ET þ C
(1)
RESULTS AND DISCUSSION
Computational procedure
Absorption properties and molecular orbitals involved
All the calculations were performed using the hybrid functional
B3LYP from the Gaussian 03 suite of program.[23] The different
structures were fully optimized in the Density Functional
Theory framework. The bond dissociation energies (BDEs)
and the triplet state energy (ET) were calculated at the
*
UB3LYP/6-31G level. The electronic absorption spectra for the
radical cations were calculated with the time-dependent density
The UV absorption properties of SiN1 and SiN2 as well as
8-hydroxyjulolidine and 3-diethylaminophenol in acetonitrile are
displayed in Fig. 1. The maximum absorption wavelengths are
summarized in Table 1. The starting amino alcohols show a broad
band between 350 and 600 nm which disappears in silylox-
yamines (Fig. 1).
**
functional theory at UB3LYP/6-311G level on the relaxed
geometries.
Figure 2 describes the highest occupied molecular orbital
(HOMOs) and lowest unoccupied molecular orbital (LUMOs)
Table 1. Absorption maxima of 8-hydroxyjulolidine, 3-diethylaminophenol, SiN1 and SiN2 in acetonitrile
SiN1
8-hydroxyjulolidine
267, 303, 462
SiN2
3-diethylaminophenol
262, 301, 526
Absorption maximum (nm)
267, 303
262, 301
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 342–350

SILYLOXYAMINES AS SOURCES OF SILYL RADICALS
Figure 2. HOMO and LUMO of SiN1 (left) and SiN2 (right). From DFT calculations
orbitals of SiN1 and SiN2; the more red-shifted absorption band
corresponding to a HOMO ! LUMO transition. The HOMO of the
silyloxyamines involves the p-system and the nitrogen lone pair.
Interestingly, the LUMO is also delocalized on the silyl ether
group. This evidences a partial charge transfer transition between
the donor (the amine) and the acceptor (the O—Si—Si group).
Silyloxyamines behave therefore as donor-p-acceptor molecules.
orbital calculations and ESR-ST results. The presence of b and c is
accounted for by a hydrogen abstraction process. The generation
of these radicals is in good agreement with the high
polymerization initiating ability observed for SiN1 and SiN2
(see below) i.e., these latter radicals being good polymerization
initiating structures.[11] In SiN1, an intramolecular process
(DH ¼ ꢀ7.1 kcal/mol; from MO calculations) can be favorable as
it involves the intermediate conformation of a six-member ring.
In SiN2, the c (aN ¼ 14.4; aH ¼ 2.4G – Table 2) amount increases
with the irradiation time in line with a bimolecular process
Photolysis experiments
It was reported[1,2–4] that silyloxy compounds can be cleaved
upon photolysis at 254 nm to give the starting alcohols with
yields ranging from 62–95%. The photolysis of SiN1 and SiN2
in degassed acetonitrile under UV irradiation (Hg–Xe lamp;
220 mW/cm2) was thus examined as shown in Fig. 3. SiN1 and
SiN2 do not generate the corresponding alcohols as the typical
absorption bands of the parent hydroxy compounds at 462
and 526 nm are not observed. This behavior can suggest the
occurrence of an Si–Si instead an O—Si bond cleavage. An
increase of a broad absorption band between 350 and 400 nm for
both silyloxyamines is also noted: it might correspond to the
formation of a photolysis product resulting from the attack of a
.
silyl radical to the aromatic ring i.e., the addition of R3Si onto
aromatic rings is well documented.[24]
ESR spin-trapping experiments
Under direct irradiation of SiN1 and SiN2, two radicals are
generated and trapped by PBN (Fig. 4). The hyperfine splitting
constants for both the nitrogen (aN) and the hydrogen (aH) of
these adducts are reported in Table 2. They are assigned
(Scheme 3) to the known trimethylsilyl radical[15] (a) and
tentatively to the carbon centered radicals (b and c). No radicals
are observed when irradiating the amino alcohols in the same
conditions. These results clearly support the Si—Si cleavage;
the O—Si bond cleavage can be probably ruled out as the
(CH3)3Si—Siꢂ(Me)2 radical is not observed here. Oxygen and
carbon centered radicals can exhibit in some case similar spectral
parameters (aN and aH);[18] however, the oxygen centered radicals
will correspond here to phenoxy derived radicals which are fairly
stable. The addition of these radicals to PBN is probably not
highly favorable. Moreover, in previous works, the parameters for
phenoxy derived radicals were found quite different to carbon
centered radicals.[19] Independently, phenoxy derived radicals
can be easily characterized by LFP experiments (intense
absorption and specific reactivity – see below the LFP data).
This absorption is clearly not observed in our LFP experiments
ruling out the O—Si cleavage in agreement with the molecular
Figure 3. Photolysis of (A) SiN1 and (B) SiN2 in acetonitrile (Hg–Xe lamp
irradiation)
J. Phys. Org. Chem. 2011, 24 342–350
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

D. L. VERSACE ET AL.
Si
Si
Si
O
O
h
ν
Si
+
a
N
N
Silyl radical
-7.1 kcal/mol
H
SiN1
Si
O
H
C
b
N
Si
Si
Si
O
O
ν
h
a
Si
+
N
N
Silyl radical
Si
Si
O
SiN2
-2.4 kcal/mol
N
Si
N
Si
O
H
C
c
Carbon-centered radical
Scheme 3.
Figure 4. ESR spin-trapping spectra using PBN. (1) SiN1 and (2) SiN2.
UV-light irradiation at l > 310 nm: (A) experimental and (B) simulated
spectra
71.3 kcal/mol
79.9 kcal/mol
69.5 kcal/mol
82.3 kcal/mol
Si
Si
Si
Si
Table 2. Hyperfine coupling constants for the radical adducts
of SiN1 and SiN2 measured by ESR spin-trapping (PBN)
O
O
aN(G)
aH(G)
N
N
SiN1
SiN2
a
b
a
c
14.8
14.4
14.8
14.4
6.3
2.5
6.3
2.4
SiN1
SiN2
Scheme 4.
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 342–350

SILYLOXYAMINES AS SOURCES OF SILYL RADICALS
(DH ¼ ꢀ2.4 kcal/mol; MO calculations). The hyperfine splitting
constants for the adduct of c are in excellent agreement with
those observed recently for aminoalkyl radicals.[20]
dissociative in MO calculations but this does not exclude the
possibility of a cleavage process.
Laser flash photolysis (LFP) investigation
Energetic considerations
The transient absorption spectra recorded immediately after the
laser flash excitation (at 266 nm) of SiN1 and SiN2 as well as
the corresponding starting amino alcohols (Fig. 5) are similar. The
8-hydroxyjulolidine and SiN1 absorption maximum wavelength
is located at 470 nm whereas it shifts to a slightly higher
wavelength (about 490 nm) for diethylamino-phenol and SiN2.
The transients decay in the 5–10 ms range. They are ascribed to
the radical cation centered on the N-atom and probably
generated through a photoionization process. Indeed, this is
consistent with the reported formation of a radical cation upon
laser excitation of various aliphatic and aromatic amines in a
freon glass,[26,27] for triethylamine, 1-azabicyclo[3.3.3]undecane,
1-azabicyclo[2.2.2]octane, or 1,4-diazabicyclo[2.2.2]octane) in
acetonitrile solutions.[28,29] Moreover, time dependent density
functional theory (TD-DFT) calculations on the possible radical
The triplet energy levels (ET) of SiN1 and SiN2 are calculated as
*
75.1 and 66 kcal/mol, respectively (at UB3LYP/6-31G level). The
different BDE associated with the cleavage of the Si—Si and
Si—O bonds in SiN1 and SiN2 are summarized in Scheme 4. The
BDEs (Si—O) are weaker than previously reported data (BDE
(Si—OR) ꢁ 110 kcal/mol with R ¼ alkyl).[25] In the present work,
these weaker BDEs (Si—O) can be ascribed to the formation of
stabilized phenoxy radicals. For BDEs (Si—Si), a good agreement
with literature data is found.[25]
Interestingly, the cleavage of the Si—Si bond is energetically
more favorable than that of the Si—O bond by about 9 and
13 kcal/mol for SiN1 and SiN2, respectively; the (Si—Si) BDEs are
lower than the (Si—O) BDEs for both compounds; the Si—Si
triplet cleavage is exothermic for SiN1 and athermic for SiN2; the
Si—O triplet cleavage is endothermic. A singlet state cleavage
should be obviously feasible in both SiN1 and SiN2 for the Si—Si
bonds. The Si—O fragmentation is not, however, experimentally
supported. The triplet state of SiN1 and SiN2 is not found
**
cation (at the B3LYP/6-311G
level; after full geometry
optimization) show here that the observed transient spectra
for 8-hydroxyjulolidine and SiN1, 3-diethylaminophenol and
SiN2 quite satisfactorily correlate with the calculated spectra
Figure 5. Transient absorption spectra of (A) 8-hydroxyjulolidine and
SiN1 and (B) 3-diethylamino-phenol and SiN2 in acetonitrile
(lexc ¼ 266 nm). The oscillator strengths for the calculated UV-visible
absorption spectra of the corresponding cations are given (sticks: see text)
Figure 6. Photopolymerization profiles of TMPTA in the presence of
(A) SiN1 (1), 8-hydroxyjulolidine (2) and SiN2 (3) in laminated conditions
(1% w/w) and (B) SiN1 (1), 8-hydroxyjulolidine (2) in aerated conditions
(1% w/w). Xe–Hg lamp; lamp intensity ¼ 44 mW/cm2
J. Phys. Org. Chem. 2011, 24 342–350
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

D. L. VERSACE ET AL.
Table 3. Radical photopolymerization of TMPTA in the presence of different photoinitiating systems: polymerization rate (Rp/[M0]
(sꢀ1)) and final conversion after 400 s of irradiation
In laminate
Final conversion (%)
Under air
Final conversion (%)
Photoinitiators
Rp/[M0] ꢃ 100
Rp/[M0] ꢃ 100
8-hydroxyjulolidine
SiN1
SiN2
62
72
68
1.4
4.1
2.2
—
17
—
—
0.2
—
(—): no photopolymerization.
(vertical bars in Fig. 5). Radicals a, b, and c are not observed due to
their lack of significant absorption.[15,16,24] The phenoxy radicals
usually characterized[30] by an intense absorption at about
400 nm are not observed here in full agreement with ESR-ST
results which have ruled out an O—Si cleavage for the
investigated compounds.
followed by a proton transfer observed for amines.[15] An
aminoalkyl radical is concomitantly generated.
Free radical photopolymerization ability
Interestingly, the SiN1 and SiN2 compounds proposed as Type I
PI are liquids: this improves their solubility in monomers
compared to the starting alcohol powders. The photopolymer-
ization profiles of TMPTA in laminated (to avoid oxygen
regeneration) and aerated conditions under a polychromatic
irradiation (Xe–Hg lamp) are shown in Fig. 6. The polymerization
rates Rp and the final conversions after 400 s of irradiation are
given in Table 3. In laminate, SiN1 and SiN2 are better PIs than
8-hydroxyjulolidine where the initiating radicals are presumably
formed through unclear by-side interactions of this amino
alcohol with the monomer. Due to their quite low absorption
The formation of the radical cation can be assumed as a
competitive pathway to the Si—Si cleavage process since the
absorption corresponding to this species is very similar for both
SiN1/8-hydroxyjulolidine and SiN2/diethylamino-phenol eviden-
cing a weak influence of the silyl moiety on this transient.
The Si—Si cleavage can result from the significant electron
density transfer from the donor (julolidine for example) to the
acceptor (silyl ether group) observed in the LUMOs. A similar
behavior has been pointed out[31] in a donor–acceptor molecule
(3,5-dimethyl-4-(9-anthracenyl)julolidine) where an intramolecu-
lar charge transfer occurs from julolidine to anthracene.
When exciting BP in the presence of SiN2, the BP ketyl radical
is detected (ketyl radical quantum yield ¼ 0.8; the procedure to
measure this parameter is presented in detail in12). The BP/SiN2
interaction rate constant is measured as 3.1 ꢃ 109 Mꢀ1 sꢀ1. The
oxidation potentials of EDB and SiN2 are 1.08 and 0.77 V,
respectively. Using a reduction potential of ꢀ1.79 eV and a triplet
state energy of 2.98 eV for BP, exothermic electron-transfer
reactions (DG ¼ ꢀ0.11 and ꢀ0.42 eV for BP/EDB and BP/SiN2,
respectively) are calculated from the Rehm–Weller equation
(Eqn 1). These results are in favor of a usual electron transfer
Figure 8. Photopolymerization profiles of TMPTA in the presence of
(1) BP (1 wt%)/EDB (3 wt%) and (2) BP (1 wt%)/SiN2 (3 wt%). In laminate.
Hg–Xe lamp; polychromatic light; intensity ¼ 44 mW/cm2
Table 4. Radical photopolymerization of TMPTA in the pre-
sence of BP/SiN2 and BP/EDB: polymerization rate Rp/[M0]
(sꢀ1) and final conversion after 400 s of irradiation
In laminate
Photoinitiating
systems
Final conversion (%)
Rp/[M0] ꢃ 100
Figure 7. Photopolymerization profiles of TMPTA in the presence of
(1) 8-hydroxyjulolidine, (2) SiN1, (3) 8-hydroxyjulolidine/TTMSS (1 wt%),
and (4) SiN1/TTMSS (1 wt%). Under air. Xe lamp; l > 400 nm. Intensity ¼
60 mW/cm2
BP/EDB
BP/SiN2
71
80
16.7
23.3
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 342–350

SILYLOXYAMINES AS SOURCES OF SILYL RADICALS
(Fig. 1) for the principal emission bands of Hg–Xe lamp i.e.,
l > 300 nm, their efficiency is likely lower than that encountered
with other Type I PIs. SiN2 is better than SiN1 in agreement
with the above-mentioned energetic considerations for a more
efficient Si—Si cleavage process. Under air, as expected in such a
low viscosity matrix,[32–34] the polymerization rate dramatically
drops down. However, interestingly, SiN1 compared to
8-hydroxyjulolidine leads to a significant polymerization process
under air. This behavior is highly worthwhile since many classical
PIs do not allow initiation for these severe conditions. This
evidences the formation of silyl radicals i.e., it is well known that
these species are useful for polymerization in aerated conditions
(to overcome the oxygen inhibition).[32–34]
[6] N. Arsu, G. Hizai, Y. Yagci, Macromol. Rep. 1995, 1257–1262.
[7] I. Kminek, Y. Yagci, W. Schnabel, Polym. Bull. 1992, 29, 277–282.
[8] C. Peinado, A. Alonso, F. Catalina, W. Schnabel, J. Photochem. Photo-
biol. A 2001, 141, 85–91.
[9] J. Lalevee, X. Allonas, J. P. Fouassier, J. Org. Chem. 2007, 72,
6434–6439.
[10] J. Lalevee, M. El-Roz, X. Allonas, J. P. Fouassier, Macromolecules 2007,
40, 8527–8530.
´
[11] J. Lalevee, B. Graff, X. Allonas, J. P. Fouassier, J. Phys. Chem. A 2007,
111, 6991–6998.
´
[12] J. Lalevee, X. Allonas, J.P. Fouassier, J. Am. Chem. Soc. 2002, 124,
9613–9621.
[13] J. Lalevee, A. Dirani, M. El-Roz, X. Allonas, J. P. Fouassier, Macromol-
ecules 2008, 41, 2003–2010.
[14] J. Lalevee, M. El-Roz, X. Allonas, J. P. Fouassier, J. Polym. Sci. A Polym.
Chem. 2008, 46, 2008–2214.
Under visible light (Xe lamp; l > 400 nm), in an aerated matrix,
both the rate of polymerization and final conversion decrease
(Fig. 7 for SiN1). To speed up the polymerization, experiments
were done using SiN1 in the presence of tris(trimethylsilyl)silane
(TTMSS; 1 wt%) already shown as an excellent additive for
photopolymerization under air.[9,32–35] A remarkable enhance-
ment is noted as quite good photopolymerization profiles and
high conversion (ꢁ60%) were obtained (Fig. 7). The better
photoinitiating ability of SiN1 in the presence of TTMSS is full
agreement with previous data on other PI. The ability of TTMSS
[15] J. Lalevee, N. Blanchard, M. El-Roz, B. Graff, X. Allonas, J. P. Fouassier,
Macromolecules 2008, 41, 4180–4186.
[16] J. Lalevee, N. Blanchard, A. C. Chany, M. El-Roz, R. Souane, B. Graff,
X. Allonas, J. P. Fouassier, Macromolecules 2009, 42, 6031–6037.
[17] J. Lalevee, X. Allonas, S. Jradi, J. P. Fouassier, Macromolecules 2006, 39,
1872–1879.
[18] A. Alberti, in Magnetic Properties of Free Radicals: Nitroxide Radicals
and Nitroxide Based High-Spin Systems, Landolt-Bornstein New
Series, Group II, Vol. 26 D (Ed.: H. Fischer,), Springer-Verlag, Berlin,
2005.
[19] C. M. Camaggi, M. J. Perkins, J. Chem. Soc. Perkin Trans. 2 1972, 5,
507–509.
´
to overcome the oxygen inhibition in FRP process is already
[32–34]
[20] J. Lalevee, D. Gigmes, D. Bertin, B. Graff, X. Allonas, J. P. Fouassier,
Chem. Phys. Lett. 2007, 438, 346–350.
described in References
.
[21] D. R. Duling, J. Magn. Reson. B 1994, 104, 105–110.
[22] D. Rehm, A. Weller, Isr. J. Chem. 1970, 8, 256–267.
[23] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, V. G. Zakrzewski, A. Montgomery, R. E. Stratmann, Jr, 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 03, Revision B.02
(2003), Pittsburgh, PA.
[24] C. Chatgilialoglu, Organosilanes in Radical Chemistry, John Wiley &
Sons, New York, 2004.
[25] R. Walsh, Acc. Chem. Res. 1981, 14, 246–252.
[26] T. Shida, Y. Nosaka, T. Kato, J. Phys. Chem. 1978, 82, 695–698.
[27] T. Shida, Electronic Absorption Spectra of Radical Ions, Elsevier Science
Publishers, Amsterdam, New York, 1988.
[28] A. M. Halpern, D. A. Forsyth, M. Nosowitz, J. Phys. Chem. 1986, 90,
2677–2679.
[29] J. C. Scaiano, L. C. Stewart, P. Livant, A. W. Majors, Can. J. Chem. 1984,
62, 1339–1343.
Based on the LFP experiments, silyloxyamines should also work
as Type II PIs. This is exemplified (Fig. 8) by the BP/SiN2 system in
laminated conditions using a polychromatic light exposure
(Xe–Hg lamp). Compared to the BP/EDB reference system, a
remarkable increase of the polymerization rate (Rp) is noted
(Table 4). The direct light absorption by SiN2 leading to a silyl
radical can contribute to the Rp enhancement.
CONCLUSION
In the present paper, the photolysis of two new silyloxyamines
derived from pentamethyldisilane is investigated. Compared to
usual super silyl ethers, they cannot be deprotected upon
photolysis as they do not yield the starting alcohol (the Si—Si
bond is weaker than the Si—O bond). On the opposite, the
generation of the silyl radicals allows the use of these compounds
as PI. Some modifications of the chromophoric group (for
example by introducing more delocalized p systems) should
allow tuning of the absorption towards the visible due to the
charge transfer character of the HOMO–LUMO transition,
enhance the light absorption properties and, hence, the
polymerization ability of these compounds: this new concept
opens up a new way for the design of efficient visible-near UV
absorbing PIs in the radiation curing area.[36–40] This concept
might also be used in nonlinear optics or two-photon absorption
applications.[41–43]
[30] P. K. Das, M. V. Encinas, S. Steenken, J. C. Scaiano, J. Am. Chem. Soc.
1981, 103, 4162–4169.
[31] J. V. Lockard, A. B. Ricks, D. T. Co, M. R. Wasielewski, J. Phys. Chem. Lett.
2010, 1, 215–218.
´
[32] M. El-Roz, J. Lalevee, X. Allonas, J. P. Fouassier, Macromolecules 2009,
42, 8725–8732.
[33] M.A. Tehfe, J. Lalevee, D. Gigmes, J.P. Fouassier, Macromolecules 2010,
´
43, 1364–1370.
[34] M.A. Tehfe, J. Lalevee, D. Gigmes, J.P. Fouassier, J. Polym. Sci. A Polym.
´
Chem. 2010, 48, 1830–1837.
[35] J. Lalevee, N. Blanchard, B. Graff, X. Allonas, J. P. Fouassier, J. Orga-
nomet. Chem. 2008, 693, 3643–3649.
REFERENCES
[36] J. P. Fouassier, Photoinitiation, Photopolymerization and Photocuring:
Fundamental and Applications, Hanser Publishers, New-York, 1995.
[37] J. P. Fouassier, (Ed.). Photochemistry and UV Curing, Research Signpost,
Trivandrum, India, 2006.
[38] J. P. Fouassier, X. Allonas, (Eds). Basics and Applications of Photo-
polymerization Reactions, Research Signpost, Trivandrum, (in press).
[39] R. S. Davidson, Exploring the Science, Technology and Applications of
UV and EB Curing, SITA Technology Ltd., London, 1999.
[1] M. A. Brook, C. Gottardo, S. Balduzzi, M. Mohamed, Tetrahedron Lett.
1997, 38, 6997–7000.
[2] M. A. Brook, S. Balduzzi, M. Mohamed, C. Gottardo, Tetrahedron 1999,
55, 10027–10040.
[3] M. B. Boxer, H. Yamamoto, Org. Lett. 2005, 7, 3127–3128.
[4] M. B. Boxer, H. Yamamoto, J. Am. Chem. Soc. 2006, 128, 48–52.
[5] M. B. Boxer, B. J. Albert, H. Yamamoto, Aldrichimica Acta 2009, 42, 3–15.
J. Phys. Org. Chem. 2011, 24 342–350
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

D. L. VERSACE ET AL.
[40] R. Schwalm, UV Coatings: Basics, Recent Developments and New
Applications, Elsevier, Oxford UK, 2007.
[42] J.P. Malval, F. Morlet-Savary, H. Chaumeil, L. Balan, D.L. Versace, M. Jin,
A. Defoin, J. Phys. Chem. C 2009, 113, 20812–20821.
´
[41] M. Jin, J. P. Malval, D. L. Versace, F. Morlet-Savary, H. Chaumeil,
A. Defoin, X. Allonas, J. P. Fouassier, Chem. Commun. 2008,
6540–6542.
[43] X. Allonas, C. Crouxte-Barghorn, J. P. Fouassier, J. Lalevee, J. P. Malval,
F. Morlet-Savary, Lasers in Chemistry (Ed.: M. Lackner), Wiley-VCH,
Weinheim, 2008.
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 342–350