Toward nitroxide-mediated photopolymerization

Article abstract of DOI:10.1021/ma902774s

A new alkoxyamine (methyl 2-((4-benzoylphenyl)(( 1 -methoxy-2-methyl-1 -oxopropan-2-yl)oxy)amino)-2-methylpropanoate 4) bearing a chromophore group directly linked to the aminoxyl function is proposed as a photoinifeiter. This original compound decomposes under UV irradiation to generate the corresponding alkyl and nitroxide radicals. Drastic changes of the photophysical or photochemical properties of the starting chromophore are noted. Laser flash photolysis experiments showed both a singlet state cleavage and an efficient shortening of the triplet state lifetime of 4. MO calculations suggest a C-O bond homolytic dissociation under UV irradiation through both S₁ and T₁ pathways. However, an ESR study evidences that both N-O and C-O homolysis occur under UV irradiation. The efficiency of 4 as a conventional photoinitiator is close to that of 2,2′,dimethoxyphenyl acetophenone. When 4 was used as a photoiniferter in nitroxide-mediated photopolymerization (NMP2), a linear growth of the poly(n-butyl acrylate) chain, reaching 80% of conversion in < 500 s, is observed combined with a reinitiation of the photopolymerization after the end of the irradiation: this is the first report showing a at least partial NMP² process.

Full text of DOI:10.1021/ma902774s

2204 Macromolecules 2010, 43, 2204–2212
DOI: 10.1021/ma902774s
Toward Nitroxide-Mediated Photopolymerization
,†
Yohann Guillaneuf,* Denis Bertin, Didier Gigmes,* Davy-Louis Versace, Jacques Lalevee,*
†
,†
‡
,‡
ꢀ
and Jean-Pierre Fouassier‡
†
ꢀ
ꢀ
UMR 6264 Laboratoire Chimie Provence, Universite de Provence, Avenue Escadrille Normandie-Niemen,
‡
ꢀ ꢀ
Case 542, Marseille 13397, Cedex 20, France, and Departement de Photochimie Generale, CNRS,
Universiteꢀ de Haute Alsace, Ecole Nationale Supeꢀrieure de Chimie de Mulhouse -ENSCMu,
3 rue Alfred Werner, 68093 Mulhouse Cedex, France
Received December 16, 2009; Revised Manuscript Received January 28, 2010
ABSTRACT: A newalkoxyamine (methyl 2-((4-benzoylphenyl)((1-methoxy-2-methyl-1-oxopropan-2-yl)oxy)-
amino)-2-methylpropanoate 4) bearing a chromophore group directly linked to the aminoxyl function is proposed
as a photoiniferter. This original compound decomposes under UV irradiation to generate the corresponding
alkyl and nitroxide radicals. Drastic changes of the photophysical or photochemical properties of the starting
chromophore are noted. Laser flash photolysis experiments showed both a singlet state cleavage and an efficient
shortening of the triplet state lifetime of 4. MO calculations suggest a C-O bond homolytic dissociation under UV
irradiation through both S1 and T1 pathways. However, an ESR study evidences that both N-O and C-O
homolysis occur under UV irradiation. The efficiency of 4 as a conventional photoinitiator is close to that of 2,20-
dimethoxyphenyl acetophenone. When 4 was used as a photoiniferter in nitroxide-mediated photopolymerization
(NMP2), a linear growth of the poly(n-butyl acrylate) chain, reaching 80% of conversion in <500 s, is observed
combined with a reinitiation of the photopolymerization after the end of the irradiation: this is the first report
showing a at least partial NMP2 process.
Introduction
Basically, during the NMP, the initiation step occurs through
the homolysis of an alkoxyamine under thermal conditions
Photopolymerization represents a rapidly emerging field in the
material science area. The actual growth of this market is ∼7%
each year. The photochemical process is characterized by impor-
tant advantages because thin films can be polymerized very rapidly
(<1 s) with UV lamp irradiation. For polymer synthesis, coatings,
inks, photoresists, dual-cure systems, and so on, the photopolym-
erization reaction is now recognized as particularly interesting and
powerful.1 This is also an ecological alternative to thermal pro-
cesses because no volatile organic compounds (VOC) are released
in this technology. However, the main drawback of the photo-
chemical processes is related to the lack of control for the final
polymer properties and the difficulty to prepare block copolymers.
To overcome this problem and open a large range of applications
for photopolymers, Otsu et al.2 developed dithiocarbamate com-
pounds called photoiniferters (see an example in Scheme 1), which
are able to initiate and control the polymer growth. With these
systems, the synthesis of complex architecture polymers (block,
star, or amphiphilic polymers, etc.) exhibiting new properties
(optic, reinforcement, chemical compatibilization, surface modifi-
cation, etc.) becomes possible. Despite very encouraging results,
the method developed by Otsu suffers from severe limitations.3
Dithiocarbamyl radicals dimerize and initiate extra chains. Up to
now, no other efficient controlled radical photopolymerization
system has been described despite the industrial and academic
growing needs in this field. To overcome the drawbacks mentioned
above, the present article will propose an original alkoxyamine as a
representative of a new family of photoinitiator-controller for
nitroxide mediated photopolymerization (NMP2).
(Scheme 2). This homolytic dissociation leads to a transient
carbon-centered radical that acts as the initiator and to a
nitroxide radical that plays the role of the controlling agent.
The NMP process is based on the persistent radical effect, and the
propagation step is described as the reversible dissociation
between macroalkoxyamine (dormant species) and the active
species, which are the carbon-centered propagating radicals.
Owing to the stability of their alkoxyamine end group, the
dormant macromolecules can be isolated and further used as
macroinitiators under thermal condition for the polymerization
of a different monomer and therefore afford the preparation of
diblock copolymers.
Very few studies have been devoted to the photodissociation
properties of alkoxyamines, which would allow the possibility of
NMP2. Scaiano et al.5 first tested different TEMPO-based
alkoxyamines (where TEMPO is 2,2,6,6-tetramethylpiperidiny-
loxy-4-yl radical) whose homolysis was photosensitized by a
xanthone (Xa) triplet state or a pyrene (Py) excited singlet state.
The quenching of the excited states is observed, but only a few
alkyl radicals were generated (in the presence of xanthone
exclusively). Yoshida6-10 also investigated the photopolymeriza-
tion of various monomers in the presence of TEMPO derivatives
or their corresponding alkoxyamines using a photoacid generator
(bis(alkylphenyl)iodonium hexafluorophosphate (BAI)) to pro-
duce radicals. Nevertheless, this mechanism remains unclear
because the role of the BAI is not elucidated. To enhance the
efficiency of the energy transfer between the chromophore and
the C-O bond, Neckers et al.11 designed new alkoxyamines
bearing different chromophores covalently linked to the TEM-
PO-OH moiety (see, e.g., the alkoxyamine 1-3, Scheme 1). With
both the xanthone and the benzophenone moiety, a rapid
quenching of the excited state during LFP experiments was
Nitroxide radicals have been already extensively used as con-
trolling agent in nitroxide-mediated polymerization (NMP).4
*Corresponding authors. E-mail: yohann.guillaneuf@univ-provence.fr;
didier.gigmes@univ-provence.fr; jacques.lalevee@uha.fr.
pubs.acs.org/Macromolecules
Published on Web 02/10/2010
r 2010 American Chemical Society

Article
Macromolecules, Vol. 43, No. 5, 2010 2205
Scheme 1. Nitroxides and the Corresponding Alkoxyamines
Scheme 2. Nitroxide Mediated Photopolymerization (NMP2)
monomer for coating industry) and n-butyl acrylate are presented
Mechanism
hereafter.
Experimental Section
Materials. 2-Bromo-2-methyl-propionic acid methyl ester
(99þ%), CuBr (98%), Cu(0) (powder, 99þ%), N,N,N0,N0,N00-
pentamethyl-diethylenetriamine (PMDETA, 99þ%), and n-
butyl acrylate (BA, 99þ%) were purchased from Aldrich and
used as received. Solvents (benzene, diethyl ether, pentane) were
purchased from SDS and used as received. The 4-nitrosobenzo-
phenone was prepared according the procedure of Defoin.13 The
alkoxyamine 1 was prepared as described by Neckers et al.11 All
reactions were monitored by analytical thin layer chromatog-
raphy (TLC) plates and analyzed with 254 nm UV light,
phosphomolybdic acid solution, or both.
observed, but the detection of the alkyl radical was not easy. The
photopolymerization of methyl methacrylate in the presence of 2
exhibits one of the criteria of a controlled polymerization (i.e., the
Mn is linked to the amount of the initiating alkoxyamine), but the
final polymer was not photoreactive under irradiation condi-
tions, indicating a nonliving character. Similarly Scaiano et al.12
used a quinoline as chromophore (Alkoxyamine 3, Scheme 1).
The photolysis between 300 and 400 nm in acetonitrile showed a
first-order increase in the nitroxide concentration in the electron
spin resonance (ESR) cavity in agreement with the unimolecular
decomposition of the alkoxyamine. Their first attempt to photo-
polymerize styrene was encouraging, even if the polydispersity
index was quite large (i.e., PDI = 1.6). A more detailed study
with this kind of alkoxyamine is needed to conclude about the
efficiency of this compound for both the controlled and living
character of the polymerization.
Instrumentation. 1H (300.13 MHz) and 13C (75.44 MHz)
NMR spectra in CDCl3 were recorded on a Bruker AC-300
spectrometer using 5 mm o.d. tubes. Chemical shifts were
reported as δ values in parts per million (ppm) relative to
residual CHCl3 as internal reference (1H: δ7.26, 13C: δ77.23),
and coupling constants (J) were reported in hertz (Hz). Peak
multiplicities are indicated as follows: s (singlet), d (doublet), t
(triplet), and q (quartet). Hydrogen multiplicity (C, CH, CH2,
CH3) information is obtained from DEPT experiments.
Elemental analyses were performed at the Service Commun
ꢀ
de Microanalyse de l’Universite d’Aix-Marseille 3 using a
Thermo Finnigan EA 1112. The experiments were carried out
under a N2 atmosphere and are analyzed with the Eager 300
software.
ESI-MS experiments were performed with a QStar Elite mass
spectrometer (Applied Biosystems SCIEX, Concord, ON,
Canada) equipped with an electrospray ionization source oper-
ated in the positive ion mode. The capillary voltage was set at
5500 V, and the cone voltage was set at 80 V. In this hybrid
instrument, ions were measured using an orthogonal accelera-
tion time-of-flight (oa-TOF) mass analyzer. Air was used as the
nebulizing gas (25 psi). Analyst software version 2.1 was used for
instrument control, data acquisition, and data processing. Ac-
curate mass measurements were performed using two reference
ions from a poly(ethylene glycol) internal standard (m/z
367.2690 and m/z 425.3108). Standards solutions were prepared
in THF, further diluted using a methanolic solution of ammo-
nium acetate (3 mM), and finally mixed with the sample prior
to ionization. Direct sample introduction was performed at a
10 μL/min flow rate using a syringe pump.
Beause the energy transfer from the sensitizer (“antenna”) to
the C-O bond is not found to be efficient enough in these
previous works, a new approach where the chromophore group is
directly linked to the aminoxyl function will be developed here to
facilitate the homolysis of the alkoxyamine (Scheme 2). In this
way, the photochemical and photophysical properties of the
chromophore will be much more strongly affected (UV absorp-
tion spectrum, singlet and triplet state lifetimes, intersystem
crossing quantum yield, dissociation quantum yield, etc.) than
when using separated chromophore and alkoxyamine fragments.
In the present article, the photochemical behavior in NMP2 of the
original alkoxyamine 4 (Scheme 1) exhibiting such properties will
be examined. Alkoxyamines synthesis as well as their photo-
chemical properties and their use in the polymerization of
both a conventional curing system (a multifunctional acrylate

2206 Macromolecules, Vol. 43, No. 5, 2010
Guillaneuf et al.
Synthesis of Alkoxyamine Methyl 2-((4-Benzoylphenyl)((1-
methoxy-2-methyl-1-oxopropan-2-yl)oxy)amino)-2-methylpro-
panoate 4. Under an inert atmosphere, a solution of 4-nitro-
sobenzophenone (4.7 mmol, 1.0 g) and 2-bromo-2-methyl-pro-
pionic acid methyl ester (14.1 mmol, 2.57 g) in benzene (10 mL)
was added to a deoxygenated mixture of CuBr (14.1 mmol, 2.03
g), PMDETA (28.2 mmol, 4.92 g), and Cu(0) (14.1 mmol, 901.7
mg) in benzene (10 mL). After 24 h of stirring at room tempera-
ture, the mixture was filtered off on Celite, diluted with diethyl
oxide, and then washed with water to extract copper-based
compounds. The organic layer was dried over anhydrous
MgSO4. After removal of the solvent at RT under reduced
pressure, the crude product was purified by silica gel column
chromatography (pentane/ether 1:1). Alkoxyamine 4 (1.3 g,
3.14 mmol, yield = 66%) was obtained as a red oil. 1H NMR
(CDCl3, 300.13 MHz, δ): 1.24-1.48 (m, 12 H), 3.35 (s, 3 H), 3.72
(s, 3 H), 7.29-7.86 (m, 10 H). 13C NMR (CDCl3, 75.47 MHz, δ):
21.2 (CH3), 23.5 (CH3), 23.9 (CH3), 24.1 (CH3), 51.4 (-O-
CH3), 51.8 (-O-CH3), 68.4 ((COOMe)(CH3)2-C-N), 81.6
((COOMe)(CH3)2-C-O), 124.8, 128.1, 129.7, 132.1, 134.5,
137.6, 152.5 (aryl carbons), 172.7 (-CdO), 173.5 (-CdO),
195.7 (-CdO). ESI-HRMS calcd for C23H27NO6 [M þ H]þ:
414.1911; found: 414.1912. Anal. Calcd. for C23H27NO6
(413.464): C, 66.81; H, 6.58; N, 3.39. Found: C, 66.34; H,
7.04; N, 3.15.
Scheme 3. Alkoxyamine Synthesis Using a Nitroso Compound
Bruker ESP 300 ESR spectrometer at 9.5 GHz (X-band) employ-
ing 100 kHz field modulation and a microwave power of 10 mW,
equipped with a variable temperature unit Bruker ER 4111 VT.
The magnetic field strength and microwave frequency were
measured with a Bruker ER 035 M NMR gaussmeter and a
Hewlett-Packard 5350B frequency counter, respectively. The
UV irradiations were performed with a 1000-W UV Xe-Hg
Oriel lamp (Newport Corp., CA).
ESR spin trapping experiments were carried out on a X-Band
spectrometer (MS 200 Magnetech). The radicals generated
during the irradiation of the investigated compounds are
trapped by phenyl-N-tbutylnitrone (PBN). Irradiation was
carried out in the cavity (TE102) of the spectrometer with the
light of a Hg-Xe lamp (Hamamatsu, L8252, 150 W) through a
filter to cut the light below 310 nm.
Computational Procedure. Molecular orbital calculations
were carried out with the Gaussian 03 suite of programs.17
The different structures were fully optimized in the density
functional theory framework (at B3LYP/6-31G* level). The
bond dissociation energies (BDEs) were calculated as the en-
ergetic difference between the radicals and the starting mole-
cules. Both N-O and C-O BDEs were calculated.
Polymerization Experiments. The photopolymerization of
BA was carried out in a sealed glass tube (sample thickness: 1
mm) after Ar bubbling under irradiation with the UV light of a
Xe-Hg lamp (Hamamatsu, L8252, 200 W). The monomer
conversion was directly followed by real time FTIR spectrosco-
py in the near-infrared spectral region as usually done for thick
samples (4700-4800 cm-1). The number-average molecular
weights (Mn) and polydispersity index (PDI = Mw/Mn) were
determined by size exclusion chromatography (SEC) using a
Waters 515 HPLC pump equipped with 2 columns (PSS SDV
LINEAR M 5 μm) and two detectors: UV/visible (Waters 486)
and DRI (Waters 2414). Measurements were performed in
tetrahydrofuran (THF) at room temperature with a flow rate
of 1 mL min-1. Toluene was used as a flow rate marker.
Calibration was based on polystyrene standards (kit EasyCal
from Polymer Laboratories; Mn between 580 and 377 400
Results and Discussion
Alkoxyamine Synthesis. Linear aryl nitroxides (such as 5,
Scheme 1) and their corresponding alkoxyamines have a UV
absorption in the 300-400 nm range because of the deloca-
lization of the nitrogen lone pair with the phenyl moiety.
Because of its unusual photochemical properties, these latter
compounds could possibly act as a PI for NMP2. Among the
numerous nitroxides prepared in the literature, very few
examples deal with nitroxides of type 5.18-20 Usually, their
synthesis is not convenient, and most nitroxides are only
persistent,21 which make the preparation of the alkoxya-
mines very difficult. To obtain the alkoxyamines based on
nitroxides 5, we followed a one-step procedure originally
developed by Grubbs et al.22,23 This method presents the
advantage of avoiding the isolation and purification of the
nitroxide and consists only of the reaction of a nitroso
compound in the presence of 2 equiv of alkyl radical gener-
ated in situ from the corresponding alkyl halide (Scheme 3).
Reaction of the alkyl halide in the presence of the Cu(I)Br/
Cu(0)/PMDETA complex was used as a convenient source
of alkyl radicals. In this system, the first equivalent of the
generated alkyl radical reacted with the nitroso derivatives to
afford in situ the corresponding nitroxide. The latter com-
pound trapped a second alkyl radical to produce the desired
alkoxyamine. We chose the benzophenone (BP) moiety
because of its availability and its well-known efficiency as a
chromophore. The synthesis of the nitroso-based benzophe-
none was straightforward using the procedure of Defoin.13
The 2-bromo-2-methyl-propionic acid methyl ester was
chosen both to increase the steric hindrance close to the
aminoxyl function and to lower the BDE (C-O) of the
corresponding alkoxyamine, as already demonstrated with
SG1-based alkoxyamine.24,25 The reaction of the 4-nitroso-
benzophenone and 3 equiv of 2-bromo-2-methyl-propionic
acid methyl ester and 4-nitrosobenzophenone gave the
alkoxyamine 4 in 66% yield (Scheme 4) after purification.
Photochemical Properties and Excited State Processes. The
UV-visible absorption spectrum of alkoxyamine 4 in acet-
onitrile is depicted in Figure 1. As for the BP chromophore,26
g mol-1) and Mark-Houwink parameters (PS K = 11.4 ꢀ
3
10-3 mL g-1, R = 0.716; PBA K = 12.2 ꢀ 10-3 mL g-1, R =
0.70).14
For film polymerization experiments, a given photoinitiator
(PI) (1% w/w) was dissolved into a bulk oligomer/monomer
formulation based on 75/25 w/w epoxyacrylate/tripropylene-
glycoldiacrylate (Ebecryl 605 from Cytec). To get a good
reproducibility of the experiments, thin samples with low PI
optical densities were used. These experimental conditions allow
a good dissipation of the heat produced during the polymeriza-
tion reaction and avoid any internal filter effects. The laminated
films (25 μm thick) deposited on a BaF2 pellet were irradiated
with the polychromatic light of a Xe-Hg lamp (Hamamatsu,
L8252, 150 W). These experiments were carried out under
laminated conditions. The evolution of the double-bond content
was continuously followed by real-time FTIR spectroscopy
(Nexus 870, Nicolet).15 The experimental conditions were de-
scribed in detail in ref 15. 2,20-Dimethoxy-2-phenyl acetophe-
none (Irgacure 651,Ciba Spec. or Esacure KB1, Lamberti) was
used as a reference.
Laser Flash Photolysis Experiments. Nanosecond laser flash
photolysis (LFP) experiments were carried out using a Q-
switched nanosecond Nd/YAG laser (λexc = 355 nm, 9 ns
pulses; energy reduced down to 10 mJ, from Powerlite 9010
Continuum) and an analyzing system consisting of a pulsed
xenon lamp, a monochromator, a fast photomultiplier, and a
transient digitizer.16
Electron Spin Resonance Experiments. ESR spectra were
recorded at room temperature, unless otherwise specified, on a

Article
Macromolecules, Vol. 43, No. 5, 2010 2207
Figure 1. UV-visible absorption spectra of (A) 4 and (B) 1 in acetoni-
trile. (See Scheme 1 for the structures.)
Scheme 4. Synthesis of Alkoxyamine 4
it exhibits a significant π-π* absorption band at λ ≈290 nm
and a weak shoulder corresponding to a n-π* transition at λ ≈
340 nm. Compared with BP (λ (π-π*) = ∼250 nm), the
absorption is, however, red-shifted, which outlines the change
of the absorption properties induced by the alkoxyamine
substitution. Contrary to 1-3, the chromophore properties
are different between the unsubstituted (BP) and the final (4)
compound. From these results, the new proposed alkoxya-
mine absorbs in a more convenient spectral range.
Laser flash photolysis (LFP) experiments performed on 4
lead to a transient characterized by an intense absorption at
525 nm (Figure 2). This transient can be confidently ascribed
to the triplet state of this compound: (i) this species is
efficiently quenched by methylnaphthalene (a well-known
triplet energy acceptor) and (ii) the corresponding absorp-
tion spectrum is similar to that observed for the benzophe-
none moiety.26 A residual species is also observed (Figure 2)
with a lifetime >10 ms. Interestingly, the initial absorption
of this species is not affected by the oxygen despite the fact
that the triplet state is quenched. This probably demon-
strates that this transient is mainly produced from a short-life
singlet state (S1). The contribution of a minor triplet state
cleavage cannot be ruled out because the triplet state lifetime
(∼1.5 μs) is shorter than that for BP (6 μs under our
experimental conditions).
Figure 2. (A) Transient absorption spectra after laser irradiation of 4 in
acetonitrile (9, t = 0 μs; b, t = 15 μs). (B) Kinetics observed at 525 nm
under argon (1) and under air (2). (C) Comparison of kinetics observed
for benzophenone BP (under air) (1) and 4 (under argon) (2) in
acetonitrile for a similar light absorption intensity.
to oxygen, the residual transient is confidently ascribed to the
nitroxide. This supports a C-O bond cleavage from S1. For a
similar amount of light absorbed for 4 and BP, the transient
observed at 525 nm is much more intense for BP (Figure 2c).
Assuming similar absorption coefficients for both triplet
states, an intersystem crossing ISC (Φisc) of ∼0.1 can be
evaluated for 4. For BP, as usual, Φisc is assumed to be 1.0.26
This result demonstrates that the substitution strongly af-
fects the ISC process. No fluorescence is observed for 4 (Φfluo
< 10-3), confirming that the cleavage process is probably
the major deactivation pathway for S1.
Considering the residual maximum absorption at about
400-500 nm (very similar to that expected for the nitroxide:
452.4 nm at MPW1PW91/6-31G* level) and its insensitivity
Electron Spin Resonance Investigation of Alkoxyamine
Photodecomposition. The interesting photochemical proper-
ties of 4 prompted us to investigate the decomposition of the

2208 Macromolecules, Vol. 43, No. 5, 2010
Guillaneuf et al.
Table 1. ESR Hyperfine Coupling Constants (in mT) for 5 and 6 and
Relative C-O/N-O Cleavage Yields
relative %
aN
aHo
aHm
g
5
6
75
25
1.007
0.829
0.245
0.340
0.088
0.165
2.0057
2.0037
are generated upon irradiation of the alkoxyamine 4
(Scheme 5).
To confirm this assumption, spin-trapping experiments
were performed using R-phenyl-N-tert-butyl nitrone (PBN)
as alkyl and alkoxyl radicals scavenger. When 4 was irra-
diated in the presence of PBN, two different species were
observed and characterized by the following hyperfine cou-
plings for the spin adducts: aN = 1.45 mT; aH =0.27 mT and
aN=1.39 mT; aH=0.16 mT. These species can be confidently
ascribed to alkyl and alkoxyl radicals, respectively, because
of the value of the hydrogen splitting constants.27 A better
characterization by isolation of the spin adducts is impos-
sible because of the instability of the latter compounds.27
The observation of both carbon- and oxygen-centered
radicals is clearly in agreement with a competition of N-O
and C-O bond dissociation. The selectivity of the spin trap
agent toward both primary radicals (C , O ) does not allow a
Figure 3. ESR spectrum of nitroxide 5.
3
3
direct access to the relative N-O versus C-O pathways.
The calculated BDEs are in agreement with a competitive
cleavage; that is, it is found that the BDE(N-O) is 143.0 kJ/
mol and BDE(C-O) is 122.1 kJ/mol, with the difference
ΔBDE = BDE (N-O) - BDE (C-O) being quite low (20.9
kJ/mol). Interestingly, for this molecule, the BDE(O-C) is
lower than the BDE(N-O), which is in agreement with the
observed nitroxide formation. A triplet energy (ET) of 265
kJ/mol has been evaluated for 4 at the UB3LYP/6-31G*
level. This value is close to that found for BP (288 kJ/mol),
demonstrating that the corresponding excited state is prob-
ably mainly localized on the BP moiety.26 From the calcu-
lated BDEs, both N-O and C-O cleavages are found to be
exothermic because ET > BDE. The design of molecules
exhibiting a higher ΔBDE can be useful to enhance the
selectivity of the C-O bond cleavage. Cleavage from S1 is
expected to be even more exothermic because ES1 > ET.
Photopolymerization experiments. Photoinitiator Abil-
ity. The PI ability of 4 has been evaluated for an acrylate
matrix (Ebecryl 605, Cytec) (Figure 5). The behavior of 4 as a
conventional PI is worthwhile because it exhibits a PI ability
close to 2,20-dimethoxyphenyl acetophenone (Irgacure 651,
Ciba). This high photoinitiating ability of 4 will be useful for
applications requiring both high polymerization rates and a
living character for the polymer formed.
Nitroxide-Mediated Photopolymerization. As previously
noticed, there are only a few studies on the possibility of
alkoxyamines to control the polymerization of acrylates or
styrenic derivatives under UV activation.11,12 All of the
already described alkoxyamines are based on styryl-TEM-
PO-OH-based compounds on which different chromo-
phores were attached. As we were working with the
benzophenone chromophore, we prepared the alkoxyamine
1 as the representative of both the work of Neckers11 and
Scaiano12 and also as a comparative compound for the
alkoxyamine 4. The ability of the different alkoxyamines to
control the photopolymerization of n-butyl acrylate was
therefore investigated. This monomer was chosen because
it does not self-polymerize under UV irradiation under our
conditions. (See the Supporting Information.) The photo-
polymerization was carried out in a sealed glass tube (sample
thickness: 1 mm) after degazing by argon bubbling. The
experimental setup is schematically presented in Scheme 6.
Figure 4. Comparison between the ESR spectrum recorded after 1 h of
photolysis and the simulated spectrum coming from the combination of
nitroxide 5 and aminyl radical 6. (See Scheme 5.)
alkoxyamine under UV irradiation by ESR. To characterize
the expected nitroxide 5 upon irradiation of the alkoxya-
mine, we first heated 4 (Figure 3) inside the cavity of the
spectrometer. This strategy to observe the nitroxide radical
was motivated by the fact that the thermal homolysis of
alkoxyamine is a well-known property and because the
synthesis of pure nitroxide 5 is not straightforward. A
0.1 M solution of 4 in tert-butylbenzene was then heated
inside the cavity of the spectrometer at 120 °C in the
presence of oxygen as alkyl radical scavenger. The ESR
spectrum observed (Figure 3) and the hyperfine coupling
constants (aN=1.007 mT, aHo=0.245 mT, aHm=0.088 mT,
g=2.0057) are consistent with a linear aromatic nitroxide
structure and are in good agreement with the literature for
such structure.
Second, we studied by ESR the behavior of the alkoxya-
mine 4 in tert-butylbenzene in a quartz tube upon direct UV
irradiation inside the ESR cavity. The comparison of the
experimental spectra with the ESR spectrum of the nitroxide
5 highlighted the presence of 5 in combination with another
radical. The computer simulation of the experimental spec-
trum obtained after 1 h of irradiation clearly indicates the
formation of two radicals species (Figure 4). From the ESR
parameters (Table 1), it can be concluded that both the
nitroxide and a aminyl radical 6 arising from the N-O
homolytic bond and C-O bond homolysis, respectively,

Article
Macromolecules, Vol. 43, No. 5, 2010 2209
Scheme 5. Competition between C-O and N-O Bond Cleavages
Scheme 6. Experimental Setup Used for the Nitroxide Mediated Photopolymerization Experiment
This allows us to follow online the conversion during the
irradiation.
The first nitroxide-mediated photopolymerization experi-
ments were performed using the alkoxyamine 1 as a model of
the alkoxyamines previously used in the literature. The
polymerization kinetic is displayed in Figure 6. A linear
increase in the conversion was observed, and 30% of con-
version is reached in 500 s. In the case of alkoxyamine 4, a
linear increase in ln [M]0/[M] was also observed, and more
interestingly, 80% of conversion was obtained in 500 s. This
represents an important increase in the rate of polymeriza-
tion. By comparison, few studies were devoted to the n-butyl
acrylate polymerization mediated by dithiocarbamate com-
pounds,28,29 but in these works, the time to reach conversion
above 60% was usually >1 h at least.
Because the polymerization rate is higher in the case of
alkoxyamine 4 compared with other photoiniferter sys-
tems (dithiocarbamates and alkoxyamine 1) and is more
similar to a classical PI system, we have to investigate if a
reversible recombination reaction inducing the controlled
character was occurring. Therefore, we checked if the life-
time of the alkoxyamine functionality could be prolonged
in the absence of any scavenger. Photolysis experiments of
10 s in acetonitrile using a similar setup as the one used in
Figure 5. Conversion versus time curves for the photopolymerization
of Ebecryl 605. Photoinitiating system (1% w/w): (a) Irgacure 651
(Ciba) and (b) 4.
polymerization process were carried out under aerated
and inert (argon degazing) conditions (Figure 7). It is worth-
while to note that the irradiation of 10 s under aerated

2210 Macromolecules, Vol. 43, No. 5, 2010
Guillaneuf et al.
Figure 6. Kinetic plot for the photopolymerization of n-butyl acrylate
using 1 or 4 (1% w/w).
Figure 7. Photolysis of 4 in acetonitrile under air or argon. The
irradiation conditions are similar to those of the photopolymerization
experiments. A full bleaching of 4 is noted in the presence of scavenger
(air) after 10 s of irradiation (280 nm). A shoulder is observed at ∼450
nm and can be tentatively ascribed to the nitroxide.
Figure 8. (a) Mn and PDI versus conversion for the photopolymeriza-
tion of n-butyl acrylate with 4 (1% w/w targeted Mn = 41 300 g mol-1
3
at 100% conversion). (b) Typical SEC curves associated with NMP2
experiments (48% of conversion) with a detection of the polymer chains
by both RI and UV-absorption λ = 280 nm (maximal absorption for
the chromophore of 4).
conditions (which allows us to trap the carbon-centered
radicals by oxygen to avoid the recombination process and
the alkoxyamine regeneration) fully decomposes 4 (dis-
appearance of the absorption band at 280 nm), whereas
under inert conditions, only a small amount of 4 was con-
sumed. This result means that the lifetime of macroradicals is
drastically prolonged, and this enables a living process.30
Second, in a real polymerization process, the monomer acts
as a scavenger for the initiating radical, which suggests that 4
was totally consumed in 10 s and that the polymerization
process for t > 10 s is initiated by the decomposition of
macroalkoxyamines.
does not take place, and almost constant Mn values versus
conversion were also found.31
When 4 is used as initiator instead of 1, a linear evolution
of Mn with the conversion is noted (Figure 8a). Interestingly,
the amplitude of the Mn change is quite large (25 000-40 000
g mol-1). This behavior suggests a partial controlled char-
3
acter for the photopolymerization process. Moreover, the
relatively low Mn obtained compared with the one obtained
using alkoxyamine 1 is another indirect proof that alkoxya-
mine 4 strongly moderates the reactivity of the polymeriza-
tion process and allows the generation of a controlled and
living system.
The molecular weight distributions for the n-butyl acrylate
polymerization performed with 1 are similar, whatever the
conversion (Supporting Information), and the Mn values are
close to 250 000 g mol-1 with PDI close to 2.5. Such high Mn
The PDIs are found to be quite large from 2.2 for a
conversion of ∼30% (irradiation time 100 s) to 3.9 for a
conversion of 80% (irradiation time 500 s), demonstrating
that the control process is not perfect. It has to be noticed
that the experimental conditions, that is the irradiation
system (geometry, UV lamp, etc.) can be of importance for
both the kinetics and the behavior (control, livingness) of the
polymerization. As light from the source penetrates the
sample, it is absorbed by the alkoxyamine, causing a decrease
in light intensity with depth into the sample. This can cause
gradients in polymerization rate and therefore also leads to
heat and mass transfer-related effects.32,27 It also has to be
3
values compared with the theoretical ones ([1] = 1% w/w,
i.e., 48 500 g mol-1 at 100% conversion) could be explained
3
by a nonefficient photolysis of 1, leading to very small
amounts of alkyl and nitroxide radicals. The alkyl radicals
initiate the polymerization of n-butyl acrylate, a monomer
with a high propagation rate constant. At the same time, the
amount of nitroxide is not sufficient to trap the alkyl
macroradicals and stop or moderate the propagation. The
combination of these phenomena led to polymers with high
Mn, as already observed in NMP.24 In a similar way, when
using Irgacure 651, the reversible recombination obviously

Article
Macromolecules, Vol. 43, No. 5, 2010 2211
the refractive index RI signal evidence that a large part of the
chains is linked to the chromophoric group; that is, the
polymer is able to reinitiate the polymerization (Figure 8b).
The nonperfect overlapping is due to the nonperfect living
character and also to the fact that mass-sensitive (RI signal)
and number-sensitive (UV signal) detectors are used. Never-
theless, these data fully evidence a partial living character for
the formed polymers.
To further illustrate the livingness of the polymer, a
reinitiation test was performed.38 A similar experiment to
the one displayed in Figures 7 and 8 was performed. The
polymerization was carried out for 200 s (39% of conversion;
a different light intensity than Figures 7 and 8 is used), and a
poly(n-butyl acrylate) with a Mn of 64 000 g mol-1 and a
3
PDI of 3.3 was obtained. After evaporation of the residual
monomer, the polymer was again diluted by a second addi-
tion of n-butyl acrylate before a second irradiation. The
reinitiation process was observed by RT-FTIR, and 30% of
conversion was obtained in <300 s, confirming the photo-
initiation ability of the starting polymer (Figure 9a). The
molecular weight distributions MWDs before and after
reinitiation (500 s, 42% of conversion) are displayed in
Figure 9b.
The clear shift of the MWD distribution before and after
reinitiation, that is, from 64 000 to 425 000 g mol-1, evi-
3
denced a good reinitiation, even if one could notice a long tail
toward lower molar masses. Because of both the N-O versus
O-C competitive dissociation and the fact that high molar
masses polymers are obtained, the control of the second bloc
is difficult to obtain. It is not possible to check if the polymer
after reinitiation still possess the chromophoric end chain
because at such a high molecular weight, the too low con-
centration of the chromophoric group prevents its detection
by dual detection RI/UV.
Conclusion
The photolysis of alkoxyamine 4 bearing a chromophoric
group directly linked to the aminoxyl function of the nitroxide
moiety is investigated. Laser flash photolysis experiments show a
singlet state cleavage and (to a minor extent) a possible triplet
state cleavage leading to a C-O bond homolysis. An additional
N-O cleavage is supported by ESR experiments.
Using this compound as a photoiniferter for the poly-
merization of n-butyl acrylate, a linear growth of the poly-
mer chain combined with a partial living character has been
evidenced: to the best of our knowledge, this is the first exam-
ple of using an efficient photosensitive nitroxide being able to
achieve fully an NMP2 process. The control process could
presumably be optimized by using other alkoxyamines exhi-
biting a better C-O bond cleavage selectivity. This kind
of compound also appears to be a promising conventional
PI for the free radical photopolymerization of multifunctional
monomers. New compounds will be proposed in forthcoming
works.
Figure 9. (a) RT-FTIR conversion versus time plot for the reinitiation
experiment (macroinitiator: 200 s of irradiation, 39% conversion);
the excess of monomer was removed under vacuum, and the polymer-
ization kineticspresented was initiated (nBA) bythemacroalkoxyamine
formed in the first step. (b) Molecular weight distribution before
(Mn=64 000 g mol-1, PDI=3.3) and after reinitiation (Mn=425 000
3
g mol-1, PDI = 2.2).
3
noted that polyacrylates are prone in both conventional and
controlled radical polymerization processes (especially when
the conversion is high) to intermolecular polymer chain
transfer,33 which leads to an important increase in the
polydispersity index due to the long-chain branches crea-
ted.34-36 Nevertheless, in that case, the important increase in
PDI versus conversion could be likely ascribed to the N-O
bond cleavage in (macro)alkoxyamine inducing a lack of
persistent radicals. For classical photoiniferters, such a
behavior was also noted; that is, the deviation from the
origin of the Mn versus conversion plot corresponds to a
usual behavior and indicates that a part of the car-
bon-carbon termination mechanism still remains in the
early stages of the reaction, the linear trend being always
ascribed to the polymeric radical/persistent radical recombi-
nation reaction.
Acknowledgment. This work was supported by the “Agence
Nationale de la Recherche” ANR under grant NMP2 07-
JCJC0143. We would like to thank Kevin Guillaneuf for his help
in Scheme 6.
To fully evidence the living character of the formed
polymers, a characterization of the polymer chains was also
carried out by SEC experiments using both RI and UV
detection.37 For experiments similar to the one presented
in Figures 7 and 8a, the comparison of the UV signal at
280 nm (maximal absorption for the chromophore of 4) and
Supporting Information Available: Typical SEC curves
associated with NMP2 experiments (n-butyl acrylate) with a
detection of the polymer chains by RI detection, conversion
versus time plot for the photopolymerization of n-butyl acrylate
without initiator, and molecular weight distributions for the
n-butyl acrylate polymerizations initiated by 1 at different

2212 Macromolecules, Vol. 43, No. 5, 2010
Guillaneuf et al.
conversions. This material is available free of charge via the
Internet at http://pubs.acs.org.
C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98,
revision A.11; Gaussian, Inc.: Pittsburgh PA, 2001. (b) Foresman J. B.;
Frisch A. In Exploring Chemistry with Electronic Structure Methods,
2nd ed.; Gaussian. Inc.: Pittsburgh, PA, 1996.
References and Notes
(18) Das, K.; Pink, M.; Rajca, S.; Rajca, A. J. Am. Chem. Soc. 2006, 128,
5334–5335.
(19) Spagnol, G.; Shiraishi, K.; Rajca, S.; Rajca, A. Chem. Commun.
2005, 5047–5049.
(1) (a) Fouassier, J. P. Photoinitiation, Photopolymerization, and
Photocuring: Fundamental and Applications; Hanser Publishers:
New York, 1995. (b) Photochemistry and UV Curing: New Trends;
Fouassier, J. P., Ed.; Research Signpost: Trivandrum, India, 2006.
(2) Otsu, T. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2121–2136.
(3) Manga, J. D.; Polton, A.; Tardi, M.; Sigwalt, P. Polym. Int. 1998,
45, 14–21.
(4) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. Rev. 2001, 101,
3661–3688.
(5) Scaiano, J. C.; Connolly, T. J.; Mohtat, N.; Pliva, C. N. Can. J.
(20) Forrester, A. R.; Hay, J. M.; Thomson, R. H. Organic Chemistry of
Stable Free Radicals; Academic Press: New York, 1968.
(21) Calder, A.; Forrester, A. R. Chem. Commun. 1967, 682–&.
(22) Grubbs, R. B.; Wegrzyn, J. K.; Xia, Q. Chem. Commun. 2005,
80–82.
(23) Greene, A. C.; Grubbs, R. B. Macromolecules 2009, 42, 4388–4390.
(24) Chauvin, F.; Dufils, P. E.; Gigmes, D.; Guillaneuf, Y.; Marque, S.
R. A.; Tordo, P.; Bertin, D. Macromolecules 2006, 39, 5238–5250.
(25) Bertin, D.; Gigmes, D.; Marque, S. R. A.; Tordo, P. Macromole-
cules 2005, 38, 2638–2650.
(26) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochem-
istry; Marcel Dekker: New York, 1993.
(27) Rosen, G. M.; Britigan, B. E.; Halpern, H. J.; Pou, S. Free Radicals:
Biology and Detection by Spin Trapping; Oxford University Press:
New York, 1999.
Chem. 1997, 75, 92–97.
(6) Yoshida, E. Colloid Polym. Sci. 2010, 288, 7–13.
(7) Yoshida, E. Colloid Polym. Sci. 2010, 288, 73–78.
(8) Yoshida, E. Colloid Polym. Sci. 2009, 287, 1417–1424.
(9) Yoshida, E. Colloid Polym. Sci. 2009, 287, 767–772.
(10) Yoshida, E. Colloid Polym. Sci. 2008, 286, 1663–1666.
(11) Hu, S.; Malpert, J. H.; Yang, X.; Neckers, D. C. Polymer 2000, 41,
445–452.
(12) Goto, A.; Scaiano, J. C.; Maretti, L. Photochem. Photobiol. Sci.
2007, 6, 833–835.
(28) Manga, J. D.; Tardi, M.; Polton, A.; Sigwalt, P. Polym. Int. 1998,
45, 243–254.
(13) Defoin, A. Synthesis 2004, 706–710.
(14) Beuermann, S.; Paquet, D. A.; McMinn, J. H.; Hutchinson, R. A.
(29) Kilambi, H.; Reddy, S. K.; Bowman, C. N. Macromolecules 2007,
40, 6131–6135.
(30) Kothe, T.; Marque, S.; Martschke, R.; Popov, M.; Fischer, H.
Macromolecules 1996, 29, 4206–4215.
(15) Lalevee, J.; Allonas, X.; Jradi, S.; Fouassier, J. P. Macromolecules
2006, 39, 1872–1879.
(16) Lalevee, J.; Allonas, X.; Fouassier, J. P. J. Am. Chem. Soc. 2002,
124, 9613–9621.
J. Chem. Soc., Perkin Trans. 2 1998, 1553–1559.
(31) Lalevee, J.; Blanchard, N.; El-Roz, M.; Allonas, X.; Fouassier, J. P.
Macromolecules 2008, 41, 2347–2352.
(32) Goodner, M. D.; Bowman, C. N. Chem. Eng. Sci. 2002, 57, 887–
900.
(33) Junkers, T.; Barner-Kowollik, C. J. Polym. Sci., Part A: Polym.
Chem. 2008, 46, 7585–7605.
(34) Castignolles, P.; Graf, R.; Parkinson, M.; Wilhelm, M.; Gaborieau,
M. Polymer 2009, 50, 2373–2383.
(35) Gaborieau, M.; Nicolas, J.; Save, M.; Charleux, B.; Vairon, J. P.;
Gilbert, R. G.; Castignolles, P. J. Chromatogr., A 2008, 1190, 215–
223.
(36) Castignolles, P. Macromol. Rapid Commun. 2009, 30, 1995–2001.
(37) Hosseini Nejad, E.; Castignolles, P.; Gilbert, R. G.; Guillaneuf, Y.
J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 2277–2289.
(38) Guillaneuf, Y.; Gigmes, D.; Marque, S. R. A.; Astolfi, P.; Greci, L.;
Tordo, P.; Bertin, D. Macromolecules 2007, 40, 3108–3114.
(17) (a) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery,
J. A., Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; , Millam, J.
M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.;
Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.;
Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G.
A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Salvador, P.; Dannen-
berg, J. J.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M.
W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez,