SG1 nitroxide analogues: A comparative study

Article abstract of DOI:10.1071/CH10123

Due to a specific balance between steric, polar, and stabilization effects, SG1 nitroxide and its corresponding alkoxyamine BlocBuilder MA are now well recognized as two of the most potent compounds in nitroxide-mediated polymerization (NMP). In this work

Full text of DOI:10.1071/CH10123

RESEARCH FRONT
Full Paper
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2010, 63, 1237–1244
SG1 Nitroxide Analogues: a Comparative Study
Jessica Marchand,A Laurent Autissier,A Yohann Guillaneuf,A
Jean-Luc Couturier,B Didier Gigmes,A and Denis BertinA,C
AUMR 6264 Laboratoire Chimie Provence. Université de Provence. Avenue Escadrille
Normandie-Niemen, Case 542, Marseille 13397, Cedex 20, France.
BARKEMA, Centre de Recherche de Rhône Alpes, rue H. Moissan,
69493 Pierre Bénite Cedex, France.
CCorresponding author. Email: denis.bertin@univ-provence.fr
Due to a specific balance between steric, polar, and stabilization effects, SG1 nitroxide and its corresponding alkoxyamine
BlocBuilder MA are now well recognized as two of the most potent compounds in nitroxide-mediated polymerization
(NMP). In this work, alternatives to SG1, based on various aldehydes, were targeted using structure–reactivity relationships
already developed by our group. Compared with SG1, we show that the substitution of the tert-butyl group on the carbon
α to the aminoxyl function by a 2-ethylhexyl group led to a new nitroxide (ETHEXNO), which exhibited an half-life
time at 120◦C similar to SG1 and a slightly slower kd for the alkoxyamine (2–3 times lower than the SG1). The styrene
polymerization mediated by the ETHEXNO nitroxide has a similar behaviour to the one mediated by the SG1 in terms
of livingness and control but the kinetics is affected (2–3 times lower). Concerning the n-butyl acrylate polymerization,
an unexpected overheating occurred at 120◦C, which led us to perform the polymerization in toluene at 100◦C. The slow
kinetics impedes the use of this nitroxide as a good alternative to SG1 and shows that the structure of the SG1 nitroxide
is already delicately optimized and finding good alternatives is not straightforward.
Manuscript received: 16 March 2010.
Manuscript accepted: 10 June 2010.
Introduction
hybrid organic and/or inorganic systems, star polymers, poly-
mer brushes, etc. The enormous commercial potential of this
simplistic chemistry could be used in various markets including
encapsulation and release of active compounds, oil and lubri-
cant additives, dispersion stabilizers (cosmetics, inks, paints,
and mineral content), the modification of surface properties
(e.g., adhesives and coatings), the modification of mass proper-
ties (e.g., impact strength and improvement of optical qualities),
microelectronics, compatibilization, composites, etc.[18]
Despite the significant progress made in the synthesis of the
BlocBuilder MA (Scheme 1),[20] the understanding of its radical
reactivity,[21] itspost-polymerizationend-groupremoval,[22] and
its use in complex macromolecular architecture synthesis,[23]
one of our main interests is to improve its technology in terms
of fundamental knowledge related to the efficiency of this
compound.
This has led us to investigate the synthesis of new SG1-like
nitroxides and their corresponding alkoxyamines using various
aldehyde reactants that exhibit different steric properties and
polarity compared to the pivaldehyde used in the SG1 synthesis.
The choice of aldehyde is dictated using different structure–
reactivity relationships already established by our group for the
determination of the dissociation and recombination rate con-
stants kd and kc, respectively. These rate constants are known
to be the key parameters in NMP.[24,25] Once the nitroxides
and the alkoxyamines have been synthesized, they are tested
in the bulk polymerization of styrene and n-butyl acrylate and
their efficiency compared with the reference SG1 nitroxide and
BlocBuilder MA alkoxyamine.
With the development of living and/or controlled radical
polymerization (CRP) techniques,[1] such as nitroxide medi-
ated polymerization (NMP),[2–5] atom transfer radical poly-
merization (ATRP),[6–9] and reversible addition–fragmentation
chain transfer (RAFT),[10–13] the control of the structure and
architecture of vinyl polymers that exhibit specific prop-
erties is now possible in many cases. NMP has received
great interest since the pioneering work of Rizzardo and
coworkers,[4] and the subsequent development of this technique
by Georges et al.[2] has shown that well controlled and living
polystyrene can be prepared usingTEMPO (2,2,6,6-tetramethyl-
1-piperidinyloxyl) radical. This nitroxide and its derivatives
are successfully applied for the controlled polymerization of
styrenic monomers at 120◦C or higher. However, the main
drawback of TEMPO-mediated polymerization lies in the dif-
ficulty to control the acrylate derivatives.[3] A new generation
of acyclic nitroxides (among them the N-tert-butyl-2-methyl-
1-phenylpropyl nitroxide[14] (TIPNO) and N-tert-butyl-N-(1-
diethylphosphono-2,2-dimethylpropyl) nitroxide (SG1)[15,16]
bearing a β-hydrogen was subsequently designed to overcome
this difficulty.
Over the past decade,Arkema in collaboration with our group
has developed a complete controlled polymer technology, which
derives from a basic alkoxyamine trademarked as BlocBuilder
MA based on the SG1 nitroxide.[17–19] This technology enables
the design of polymers with very different architectures
and compositions such as block copolymers, functional-
ized macromolecules, graft copolymers, gradient copolymers,
)
© CSIRO 2010
10.1071/CH10123
0004-9425/10/081237

RESEARCH FRONT
1238
J. Marchand et al.
N
O
P(O)(OEt)2
kd
kc
ϩ
N
O
P(O)(OEt)2
COOH
COOH
BlocBuilder
SG1
1-Carboxy-1-methylethyl radical
Scheme 1. Decomposition of the BlocBuilder MA alkoxyamine.
Table 1. Electrical Hammett constants σi and modified Taft steric
constants ri used in Eqns 1–3
Results and Discussion
Design of New Nitroxides and Alkoxyamines
The main equilibrium between dormant and active species is
the key step in the NMP process. If the equilibrium constant
K = kd/kc (with kd the dissociation rate constant and kc the
recombination rate constant) is too high, the nitroxide acts only
as a spectator and the macroradical concentration is close to
the classical steady-state value. If the K value is too low, the
polymerization is inhibited.[24,25] The needs of estimated val-
ues of kd for designing new polymerization experiments led our
group to develop a structure–reactivity relationship based on
the 1-phenyl-ethyl alkyl radical and various nitroxides. Marque
et al.[26,27] developed a linear multiparameter relationship based
on the polar Hammett constant σL and on the Taft-type steric
constant Es to take into account the two main parameters, that is
the polarity and the bulkiness of the two alkyl groups attached
to the aminoxyl function. For non-cyclic nitroxides, steric hin-
drance around the nitroxide moiety is too high. As such, the
molecule adopts a new conformation to relieve the steric strain
and, therefore, the steric effect cannot be determined as the sum
oftheTaft-typestericconstantEs ofthetwoalkylgroupsattached
to the aminoxyl function. A levelled steric effect should then
be used.[28] The equations used to estimate the kd value are
summarized below:
R
σi
ri
Me
tBu
H
P(O)(OEt)2
CCl3
iPr
−0.01
−0.01
0
0.32
0.36
0.01
0.01
0
−2.46
0.32
−1.22
−2.98
−1.08
−1.4
2-Ethylhexyl
σL and ES,tot for different nitroxides.
log kc [M−1 s−1] = 10.35( 0.11) + 0.47( 0.17)
× σL + 0.43( 0.02) × ES,tot
(3)
with ES,tot the total steric constant taking into account the two
groups attached to the nitrogen. In that calculation, only the
four α,αꢀ substituents that flank the aminoxyl function have to
be taken into account for the steric demand. One methyl of the
tert-butyl group attached to the nitrogen and the β-hydrogen are
located in the same plane as the aminoxyl function and does not
interfere in the recombination.
The equations described above were then used to deter-
mine the kd and kc values of different nitroxides bearing a R1
group incorporated from several aldehyde reactants (Tables 1
and 2).These parameters were compared with the reference SG1
(Table 2).
Among the various aldehydes, isobutyraldehyde, 2,2,2-
trichloracetaldehyde or chloral, and 2-ethylhexanal were found
to be an interesting alternative to pivalaldehyde.These aldehydes
are commercially available and give similar kd and kc values.The
nitroxide based on chloral has a similar steric hindrance but a
strongly different polarity, whereas the opposite occurs with both
isobutyraldehyde and 2-ethylhexanal.
To compare theoretically the behaviour of all these nitro-
xides, we used the phase diagram approach developed by
Fischer.[24,35,36] The double logarithmic plot of kd versus kc is
depicted in Fig. 1. This diagram does not take into account the
penultimate effect,[31] which is known to decrease generally the
kc value by one order of magnitude.
log kd [s−1] = −5.88( 0.16) − 3.07( 0.28)
× σL − 0.88( 0.04) × ES
(1)
ꢀ
6
with σL = i=1 σL(Ri) and σL(Ri) the Hammett constant σL of
the six groups linked to the two carbons adjacent to the aminoxyl
function and
ES = −2.104 + 3.429 × r1 + 1.978 × r2 + 0.649 × r3 (2)
with ri the individual steric constant of the C(R1)(R2)(R3) group
bearing by the nitrogen and which is the part carrying the hydro-
gen atom in the β position and/or the largest group. It has to
be mentioned that the larger the Ri group, the smaller ri, thus
r1 corresponds to the small size group, r2 to the medium size
group, and r3 to the large size group.
The dissociation rate constant values could lie in the range of
10 and 10−10 s−1 between 20 and 120◦C, and then is really the
main parameter to be tuned.[29] Nevertheless, the recombination
rate constant kc, whose value could only vary between 105 and
108 L mol−1 s−1 over the same temperature range,[30,31] could
also lead to drastically different behaviours for the control and
the livingness of the polymerization.[32] The estimation of the
kc value by such structure–reactivity relationships should then
bring more insight. This study was also performed by Marque
and coworkers,[27,33,34] who determine Eqn 3[29] that links the
kc value between styryl radicals and the molecular descriptors
The close theoretical kc and kd values for the four nitroxides
(the points have a close localization in the phase diagram) shows
that their efficiency in the NMP of styrene, and hopefully of
n-butyl acrylate, should be similar.
Synthesis of New Nitroxides and Alkoxyamines
The β-phosphorylated nitroxides were prepared as previously
described[16,37] and the synthetic route is shown in Scheme 2.

RESEARCH FRONT
SG1 Nitroxide Analogues
1239
Table 2. Total electrical Hammett constants σL and modified Taft steric constants ES and EꢀS used for the
calculation of the dissociation and recombination rate constants kd and kc, respectively
Nitroxide
σL
ES
Theoretical kd (120◦C) [s−1
]
ESꢀ
Theoretical kc [M−1 s−1
]
3a (SG1)
0.28
0.3
0.65
0.3
−5.016
−3.935
−5.354
−4.329
4.7 × 10−3
4.6 × 10−4
6.8 × 10−4
1.0 × 10−3
−7.476
−6.395
−7.814
−6.788
8.9 × 106
2.6 × 107
9.5 × 106
1.8 × 107
3b (iPr-SG1)
3c (Chloral-SG1)
3d (ETHEXNO)
labile alkoxyamine[38] and for comparison with BlocBuilder
MA. The ETHEXNO-based alkoxyamines have been synthe-
sized using the atom transfer radical addition method developed
by Matyjaszewski et al.[39] In this method, the radical R• is gener-
ated by copper(i) reduction of the corresponding organic halide
RX. Thus, alkoxyamines arise from the recombination of the
nitroxide with this alkyl radical (Scheme 3).
All the alkoxyamines were obtained as mixtures of
diastereoisomers and were not separated since their reactivities
are similar (see below). The alkoxyamine 4a was obtained as
an oil in 40% yield. The alkoxyamine 4b was precipitated in
pentane and was obtained in 65% yield as a white powder.
1
0
3a
3b
3c
3d
Ϫ1
Ϫ2
Ϫ3
Ϫ4
Ϫ5
Polymerization time
above 20 h
Electron Paramagnetic Resonance (EPR) Studies
4
5
6
7
8
9
The nitroxides were studied by EPR and the hyperfine splitting
coupling constants are summarized in Table 3.
Recombination rate constant log kc [MϪ1 sϪ1
]
The hyperfine splitting coupling constants for the phospho-
rous and nitrogen atom of nitroxides 3b (iPr-SG1) and 3d
(ETHEXNO) are very close to the one of nitroxide 3a or SG1
(aP = 46.2 G, aN = 13.6 G). The hyperfine splitting coupling
constants of the nitroxide 3c or Chloral-SG1 are more dissimilar
(aP = 34.4 G, aN = 12.5 G) since the high electron-withdrawing
group trichloromethyl induced a change in the electronic density
of the aminoxyl part.
Concerning the nitroxide 3b, a hyperfine splitting coupling
constant to the β-hydrogen atom was resolved. This means that
the dihedral angle between the planes N–Cα–Hβ and Cα–N–
2pz is less close to 90◦ compared with SG1. Nitroxides bearing
hydrogen atoms on the carbon adjacent to the aminoxyl function
usually decay through a bimolecular process to yield a nitrone
and a hydroxylamine. Because of its specific conformation, the
β-hydrogen of the SG1 is eclipsed by the bulky tert-butyl alkyl
group. This conformation explains its stability, and prevents the
disproportionation of the SG1. In the case of nitroxide 3b, the
differenceinstericstrainleadstoahigherinstabilityandprevents
its use in NMP. The nitroxide 3d (ETHEXNO) was found to be
the nitroxide most similar to SG1.
The thermal stability of the ETHEXNO nitroxide was deter-
mined by monitoring the intensity of its EPR spectrum as a
function of time at a given temperature (i.e., 120◦C). This stabil-
ity was similar or even better than that of SG1 since the half-life
time at 120◦C under argon atmosphere was found to be 19 h
(15 h for SG1) and 21 h under aerated conditions (20 h for SG1).
The experimental rate constants kd were measured using
either the plateau of the increasing ESR nitroxide signal or
the decay of the 31P NMR signal, in the presence of phenyl-
hydrazine as an alkyl radical scavenger. The advantage to
perform the measurement by 31P NMR is that the dissociation
rate constants of different diastereoisomers could be determined
separately.[40] For each alkoxyamine three runs were carried
out in tert-butylbenzene as solvent.The average kd value and the
Fig. 1. Phase diagram for a the styrene polymerization at 120◦C
(kp = 2050 M−1 s−1
;
kt = 1.5 × 108 M−1 s−1
;
[Alkoxyamine]0 = 0.05 M;
Limit criteria Φlim = 0.3, δlim = 0.3, and t90 = 20 h).
First the imines 1b–d were prepared from tert-butylamine
and the desired aldehyde by reductive condensation in one high-
yielding step. In situ addition of diethyl phosphate, followed by
boron trifluoride etherate, catalyzed the reaction under ambient
atmosphere to give the corresponding aminophosphonate 2b–d.
The aminophosphonates were then oxidized in the presence of
meta-chloroperbenzoic acid (mCPBA) to synthesize the desired
nitroxides 3b–d.
The nitroxide N-tert-butyl-N-[2-ethyl-(1-diethoxyphosphoryl)
hexyl] nitroxide 3d or ETHEXNO was obtained as an orange oil
with 60% overall yield after purification. Unlike this nitroxide,
thenitroxideN-tert-butyl-N-[1-(diethoxyphosphoryl)-2-methyl-
propyl] nitroxide 3b or iPr-SG1 and the nitroxide N-tert-butyl-
N-[1-(diethoxyphosphoryl)-2,2,2-trichloroethyl] nitroxide 3c or
Chloral-SG1 were obtained with low (20%) and very low (<5%)
yield, respectively. A deeper investigation of the synthesis of
nitroxide 3b led to the conclusion that this nitroxide is suroxi-
dized to the unstable N-oxo-ammonium derivative. Concerning
the Chloral-SG1 3c, the presence of two electron-withdrawing
groups on the same carbon α to the nitrogen atom drastically
decreases the nucleophilicity of the N atom and probably reduces
its ability to react readily with peroxides or other oxidizing
agents.
To evaluate the possibilities offered by the ETHEXNO
nitroxide, we prepared two different alkoxyamines. First we
synthesized one with a phenyl-ethyl alkyl moiety to mimic the
polystyrene chain end and to determine the kinetic rate con-
stants used in the propagation step. Second we prepared one
with a 1-carboxy-1-methylethyl alkyl moiety to obtain a highly

RESEARCH FRONT
1240
J. Marchand et al.
EtO
OEt
HP(O)(OEt)2
BF3OEt2
O
P
MS 4A
R
O
ϩ
NH2
R
N
R
N
H
2a–d
1a–d
R: a ϭ tBu
b ϭ iPr
c ϭ CCl3
d ϭ 2-ethylhexyl
EtO
EtO
OEt
OEt
O
P
O
P
mCPBA
R
N
H
R
N
O
CH2Cl2
2a–d
3a–d
Scheme 2. Synthesis of β-phosphorylated nitroxides.
EtO
OEt
EtO
OEt
N
O
P
O
P
Cu/CuBr/PMDETA
N
O
R
Br
ϩ
CH2Cl2, RT, 2 h
O
R
3d
R ϭ
R ϭ
4
5
R ϭ
6
7
O
O
R ϭ
OH
OH
Scheme 3. Synthesis of β-phosphorylated nitroxide-based alkoxyamines.
coworkers.[42] These values are quite similar to the ones deter-
mined theoretically (see Table 3) and, therefore, validate the
structure–reactivity relationships already developed. The com-
parison of these values with the literature data show that they are
quite similar to those of the TIPNO nitroxide[43] (Table 4) and
lower than those of SG1. This is attributable to a slower steric
hindrance induced by the 2-ethyl-hexyl group compared with the
tert-butyl group. Nevertheless the decomposition kinetics simi-
lar to the TIPNO nitroxide[43] let us consider a good efficiency
for the ETHEXNO as a control agent in NMP.
Table 3. Hyperfine splitting coupling constants of the different
nitroxides in tert-butylbenzene (5 × 10−4 M)
Nitroxide
Hyperfine splitting coupling constants
3a (SG1)
aN 13.6 G, aP 46.2 G
3b (iPr-SG1)
3c (Chloral-SG1)
3d (ETHEXNO)
aN 13.9 G, aP 49.6 G, aH 1.2 G
aN 12.5 G, aP 34.4 G
aN 13.8 G, aP 46.8 G
Table 4. Dissociation rate constant kd and activation energy Ea for the
Styrene and n-Butyl Acrylate Polymerization
TIPNO, SG1, and ETHEXNO based alkoxyamines
The styrene polymerizations were carried out in bulk at 120◦C.
To compare the behaviour of the nitroxide 3d (ETHEXNO)
and SG1, the alkoxyamines bearing either the 1-phenylethyl
group or the 1-carboxy-1-methylethyl moiety were used. In all
cases, a linear or a quasi linear increase of the ln [M]0/[M] was
observed (Fig. 2a) even if the kinetics for the alkoxyamines bear-
ing the ETHEXNO nitroxide were close to two times slower
than the ones obtained with the SG1-based alkoxyamines. We
have already shown that for the BlocBuilder MA alkoxyamine
and for the styryl-SG1 alkoxyamine the kinetics do not follow
a t2/3 law as the acrylate-based SG1 alkoxyamines but a first
order kinetics because of the important release of nitroxide dur-
ing the quick establishment of the persistent radical effect.[38]
It is the reason why we plot ln [M]0/[M] versus time on Fig. 2.
A closer look of the kinetics (ln(ln([M]0/[M]) = f (ln t))) of the
polymerization initiated with alkoxyamines 6 and 7 showed that
polymerization initiated by 7 followed first order kinetics and the
one initiated by 6 a t2/3 law.The dissociation rate constant of 6 is
low enough to obtain a slow establishment of the PRE as already
observed with the Monams alkoxyamine in ref. [38].The tertiary
1-carboxy-1-methylethyl radical in the case of alkoxyamine 7
increases the rate of dissociation and this compound behaves
Alkoxyamine
kd (120◦C) [s−1
]
Ea [kJ mol−1
]
Styryl-SG1
6
BlocBuilder MA
7
Styryl-TIPNO
5.5 × 10−3 [41]
1.3 × 10−3
0.3[38]
124.5[41]
129.9
111.7[38]
116.2
8.5 × 10−2
3.3 × 10−3 [43]
129.6[43]
activation energy are reported in Table 4. The activation energy
Ea was estimated using A = 2.4 × 1014
s
−1. The estimated Ea
value given inTable 4 corresponds to Ea averaged over the inves-
tigated temperature range. Individual values differed by less than
2 kJ mol−1 from the average value presented in Table 4.
The activation energies for the alkoxyamines 4a and 4b are
129.9 and 116.2 kJ mol−1, respectively. In both cases the dif-
ferent diastereoisomers have the same decomposition kinetics.
This is similar to the case of the styryl-SG1 alkoxyamine[41]
and different from the alkoxyamines developed by Catala and

RESEARCH FRONT
SG1 Nitroxide Analogues
1241
(a)
to the persistent radical effect, this enabled us to determine
the K = kd/kc value (1.0 × 10−9 M−1). Using this value and
considering no chain length effect for kd (kd for a macromolec-
ular species is equal to that for the 1-phenyl-ethyl moiety) the
experimental kc value could be evaluated for the macromolecu-
lar species. This value is equal to 1.3 × 106 M−1 s−1 at 120◦C,
which is two- to three-fold higher value than that for the SG1
nitroxide. This value is also in good agreement with the theo-
retical one (see Table 2), with a ratio of kc between ETHEXNO
and SG1 of two.
Concerning the evolution of the number-average molecular
weight (Mn) versus conversion, for the four alkoxyamines the
experimental and theoretical Mn are in good agreement (Fig. 2b)
and the obtained PDI are low (<1.2 after 40% conversion). The
livingness of the polymers was also checked since this charac-
teristic is of paramount importance to prepare block copolymers
essential for the preparation of nanostructured materials and
telechelic polymers. The living fraction was determined by ESR
as already performed in ref. [38]. Typically a 10−4 M solution
of the polymer in tert-butylbenzene was heated at 120◦C for 2 h
in the presence of an alkyl radical scavenger and the amount of
released radical was determined by comparison with nitroxide
standards. For each experiment, the livingness was unambigu-
ously above 70% and the evolution of the living fraction versus
conversion is similar for the ETHEXNO and the SG1 nitroxide
(Fig. 2c).
All these results show that the control and livingness of the
styrene polymerization mediated by the ETHEXNO nitroxide
are similar to the ones obtained with the SG1 but the kinetics
are affected (2–3 times lower). This prompted us to evaluate the
potentialofthisnitroxidetocontrolthepolymerizationof n-butyl
acrylate. Indeed these all-acrylic copolymers are of wide interest
since they behave as new tough thermoplastic materials.[18]
The direct extrapolation of the experimental procedure (bulk,
120◦C) developed for the BlocBuilder MA alkoxyamine to the
new alkoxyamine 7 led to overheating during the polymerization
process and this induced a non-controlled polymerization even
if preliminary tests showed that the obtained polymer was liv-
ing. In order to avoid this overheating, the reaction temperature
was decreased to 100◦C and the polymerization was also per-
formed in 50 wt-% toluene. Under these conditions the kinetics
is controlled, the evolution of Mn versus conversion is linear and
close to the theoretical Mn and the livingness of the polymer is
similar to the one obtained with SG1 (Fig. 3). Nevertheless the
time needed to reach 60% of conversion is close to 13 h and the
PDIs are higher than those obtained with the SG1 nitroxide (1.4
instead of 1.2). Work is in progress to understand this unusual
behaviour. This result highlights again the dramatic influence of
the steric hindrance in the quality of control and livingness for
the NMP of acrylate derivatives.
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Styryl-SG1
BlocBuilder
6
7
0
50
100 150 200 250 300 350 400
Time [min]
(b)
16000
14000
12000
10000
8000
6000
4000
2000
0
3.0
M
n PDI
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
BlocBuilder
Styryl-SG1
7
6
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Conversion
(c)
100
80
60
40
ETHEXNO
SG1
20
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Conversion
Fig. 2. (a)Experimentalkineticplotofln([M]0/[M])versustimeforstyrene
bulk polymerization at 120◦C initiated with various alkoxyamines attached
to the SG1 and ETHEXNO nitroxide. (b) Evolution of number-average molar
mass (Mn, full symbols) and polydispersity index (PDI, empty symbols)
versus conversion for bulk styrene polymerization at 120◦C initiated with
various alkoxyamines (targeted Mn at 100% conversion = 20000 g mol−1).
The solid line corresponds to the theoretical Mn. (c) Evolution of the living
fraction versus conversion for the bulk styrene polymerization initiated by
the BlocBuilder MA and alkoxyamine 7.
Conclusions
Alternatives to the SG1nitroxide based on various aldehydes
were considered using structure–reactivity relationships to select
the most appropriate candidates. The substitution of the tert-
butyl group on the carbon α to the aminoxyl function by a
2-ethylhexyl group led to a new nitroxide (ETHEXNO) with
a close conformation as evidenced by ESR. The synthesis of
both the nitroxide and the corresponding alkoxyamines were
performed using the procedure already described for the SG1
and the compounds were obtained in good yields. The dissoci-
ation rate constant kd for ETHEXNO-based alkoxyamine was
like highly labile SG1-based alkoxyamines as already observed
for the Blocbuilder MA alkoxyamine.[38]
As the kinetic plot of the polymerization initiated by 6 is lin-
ear versus t1/3, thus exhibiting the typical behaviour attributable

RESEARCH FRONT
1242
J. Marchand et al.
(a)
is similar to that mediated by SG1 in terms of livingness but the
control (PDI close to 1.4) is lower and above all the kinetics is
strongly affected (60% conversion in 13 h).This impedes the use
of this nitroxide as a good alternative to SG1. This study also
highlights that the structure of the SG1 nitroxide is already a
delicate compromise and it is very difficult to find other stable
and efficient nitroxides to control the polymerization of styrenic
and acrylic derivatives.
0.7
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Experimental
Materials
All reagents were purchased from Aldrich at the highest purity
level available and used without further purification unless
otherwise stated. The SG1 nitroxide and the BlocBuilder MA
alkoxyamine were kindly provided byArkema. Silica gel for col-
umn chromatography was Merck silica gel 60.All reactions were
monitored by analytical TLC plates and analyzed with 254 nm
UV light and/or phosphomolybdic acid solution.
0
100
200
300
400
500
600
Time [min]
(b)
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
12000
10000
8000
6000
4000
2000
0
Analytical Techniques
1H, 13C, and 31P NMR spectra were recorded in CDCl3 on a
Bruker AC-300 spectrometer using 5 mm o.d. tubes. Conver-
sion was determined by 1H NMR spectroscopy. Mass spec-
trometry was performed on a 3200 QTrap machine (Applied
Biosystems Sciex) equipped with an atmospheric ioniza-
tion source. The Mn and PDI of the samples were deter-
mined by size exclusion chromatography using a Waters
515 HPLC pump equipped with three Styragel columns
(HR 3 (4.6 mm × 300 mm, separation between 500 and
30000 g mol−1), HR 4 (4.6 mm × 300 mm, separation between
5000 and 600000 g mol−1), and HR 5 (4.6 mm × 300 mm,
separation between 2000 and 4 × 106 g mol−1)), and two detec-
tors: UV/visible (Waters 486) and RI (Waters 2414). Mea-
surements were performed in tetrahydrofuran (THF) at room
temperature, with a flow rate of 1 mL min−1. Calibration was
based on polystyrene standards (kit EasyCal Polymer Labo-
ratories, Mn = 1180 to 377400 g mol−1). Data acquisition and
processing were performed with Millenium 32 Waters soft-
ware. ESR experiments were carried out on a Bruker EMX 300
spectrometer.
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Conversion
(c)
100
90
80
70
60
50
40
30
20
10
0
ETHEXNO
SG1
0.0 0.1
0.2
0.3
Conversion
0.4
0.5 0.6
N-(2-Methylpropan)-2-trichloromethyl-1-imine 1c
Trichloroacetaldehyde (0.04 mol), was diluted with 10 volumes
of ether and cooled to 0◦C.Tert-butylamine (0.12 mol) was com-
bined and titanium chloride (0.022 mol) was slowly added. The
reaction mixture was strongly stirred for 2 h at room tempera-
ture. The mixture was washed with 150 mL of NaOH solution
(0.75 M).The aqueous phase was extracted three times with ether
and the combined organic layers were dried over magnesium sul-
fate. The solvent was removed under pressure to afford a yellow
oil (5.7 g, 70%). δH 1.32 (s, 9H, C–CH3), 7.67 (d, 1H, CH=N).
δC 29.5 (3C, C(CH3)3), 58.1 (C–N), 95.9 (C–Cl3), 152.4 (C=N).
Fig. 3. (a) Experimental kinetic plot conversion and ln([M]0/[M]) ver-
sus time for the n-butyl acrylate polymerization (toluene 50 wt-%) at
100◦C initiated with alkoxyamine 7. (b) Evolution of number-average molar
mass (Mn, square symbols) and polydispersity index (PDI, diamond sym-
bols) versus conversion for the n-butyl acrylate polymerization (toluene
50 wt-%) at 100◦C initiated with alkoxyamine 7 (targeted Mn at 100%
conversion = 20000 g mol−1). The solid line corresponds to the theoret-
ical Mn. (c) Evolution of the living fraction versus conversion for the
n-butyl acrylate polymerization (toluene 50 wt-%) at 100◦C initiated with
alkoxyamine 7.
N-(2-Methylpropan)-2-ethylhexyl-1-imine 1d
determined and was found to be between two and three times
lower than the SG1 analogue.
2-Ethylhexanal (0.122 mol) was cooled at 0◦C in the presence
of 4 Å molecular sieve (5 g). Tert-butylamine (0.134 mol) was
then added and the reaction mixture was stirred for 4 h at room
temperature. It was then filtered and residual tert-butylamine
was removed under reduced pressure to give a colourless oil
(21.7 g, 86%). δH 0.85 (m, 6H, CH3CH2), 1.15 (s, 9H, C–CH3),
1.20–1.55 (m, 8H, CH2), 2.11 (m, 1H, CH), 7.25 (d, 1H,
CH=N). δH 11.9 (CH3CH2), 14.3 (CH3–CH2–CH2), 23.1
The styrene polymerization mediated by the ETHEXNO
nitroxide has a similar behaviour to that mediated by SG1 in
terms of livingness and control but the kinetics is affected (2–3
times lower). Concerning the n-butyl acrylate polymerization,
an unexpected overheating occurred using the common exper-
imental procedure (bulk, 120◦C), which led us to perform the
polymerization in toluene at 100◦C. In that case the behaviour

RESEARCH FRONT
SG1 Nitroxide Analogues
1243
(CH3–CH2–CH2), 26.1 (CH2–CH2–CH2), 29.7 (CH3–CH2–
CH), 30.1(3 × CH3–C), 32.6(CH2–CH2–CH2), 47.4(CH), 56.9
(C), 163.4 (CH=N).
44.6–45.1 (CH3–CH2–O), 50.2–50.4 (CHCH2), 51.2–51.5 (C),
52.1–52.4 (CHCH), 61.6–61.7 (CH–NH), 62.4–62.5 (CH3–
CH2–O). δP 29.38 (48%), 29.53 (52%).
Diethyl [1-(tert-Butylamino)-2-methylpropyl]
Phosphonate 2b
General Procedure for Oxidation of Aminophosphonate
into Nitroxide
In a two necked round-bottom flask, isobutyraldehyde (0.5 mol)
was cooled to 10◦C under nitrogen. Tert-butylamine (0.5 mol)
was added dropwise and after addition the mixture was heated
at 35◦C for 2 h. Water was removed from the flask and 4 Å
molecular sieves were added to the solution. Diethyl phosphate
(0.75 mol) was then added and the mixture was heated at 40◦C
for 17 h. The mixture was then diluted with dichloromethane
and filtered over celite. The solution was acidified to pH 3 with
dilute chlorhydric acid (5% v/v) and extracted five times with
dichloromethane. The aqueous phase was treated with a satu-
rated NaHCO3 solution to pH 8 and residual aminophosphonate
was extracted two times with dichloromethane.The solution was
dried over magnesium sulfate and the solvent removed under
pressure to give a colourless oil (79.8 g, 60%). δP 29.9. δH 0.90
(t, 3H, JH–H 6.9, O–CH2–CH3), 0.91 (t, 3H, JH–H 6.8, O–CH2–
CH3), 0.98 (s, 9H, C–CH3), 1.20 (t, 3H, JH–H 7.1, O–CH2–CH3),
1.21 (t, 3H, JH–H 7.1, O–CH2–CH3), 1.96 (m, 1H), 2.80 (dd, 1H,
JH–P19.3, JH–H 2.9), 3.95–4.08 (m, 4H). δC 16.5 (d, 2C, JC–P 5.6,
2C, CH3CH2), 18.8 (s, 1C, CH3CH), 19.6 (d, 1C, JC–P 11.2,
CH3CH), 30.3 (s, 3C, CH3C), 31.6 (d, 1C, JC–P 6.4, CH3CH),
50.9 (d, 1C, JC–P 5.9, CH3C), 54.7 (d, 1C, JC–P 146.3, CH–P),
61.4 (d, 1C, JC–P 7.5, CH2), 62.0 (d, 1C, JC–P7.5, CH2).
Aminophosphonate (0.016 mol) was dissolved in dichloro-
methane(80 mL)andthencooledat0◦C. Meta-chloroperbenzoic
acid (0.048 mol) dissolved in dichloromethane (95 mL) was
added dropwise over a period of 2 h. The reaction mixture was
stirred overnight at room temperature. It was then filtered and
the residue was washed several times with a saturated solution
of Na2CO3 and with a solution of 5% HCl in water. The organic
phase was dried over MgSO4 and the solvent was removed under
reduced pressure to give the nitroxide.
N-tert-Butyl-N-[2-ethyl-(1-diethoxyphosphoryl-
2-methylpropyl)] Nitroxide 3b
Purification: chromatography (pentane/ethyl acetate, 1/1).
Nitroxide: orange oil (yield 26%). m/z (ESI): 336 (M+). ESR:
aN 13.9 G, aP 49.6 G, aH 1.2 G.
N-tert-Butyl-N-[2-ethyl-(1-diethoxyphosphoryl)-2,2,2-
trichloroethyl] Nitroxide 3c or Chloral SG1
Nitroxide: yield <5%. ESR: aP 34.4 G, aN 12.55 G.
N-tert-Butyl-N-[2-ethyl-(1-diethoxyphosphoryl)hexyl]
Nitroxide 3d or ETHEXNO
Purification: chromatography (pentane/ethyl acetate, 3/1).
Nitroxide: orange oil (3.2 g, yield 64%). m/z (ESI) 336 (M+).
ESR spectra: aP 46.8 G, aN 13.8 G.
Diethyl [1-(tert-Butylamino)-2-trichloroethyl]
Phosphonate 2c
Imine 1c (0.015 mol) and diethyl phosphate (0.015 mol) were
cooled at 0◦C. BF3·OEt (0.0015 mol) was added. Formation of
the aminophosphonate was monitoring byTLC.Themixture was
then diluted with dichloromethane and filtered over celite. The
solution was dried over MgSO4 and the solvent was removed
under reduced pressure to give a white powder (5.1 g, 70%).
δP 21.70. δH 1.23 (s, 9H, C–CH3), 1.38–1.40 (t, 6H, JH–H 7.1,
2 × CH3–CH2–O), 2.35 (d, 1H, JH–P 19, CH), 4.10–4.30 (m,
4H, JH–H 7.1, 2 × CH3–CH2–O). δC 16.6 (d, JC–P 9, 2C, CH3–
CH2–O), 30.7 (d, JC–P 40, 3C, C(CH3)3), 52.3 (CH3–C–N), 62.8
(d, JC–P 9, P–CH–N), 63.8 (d, JC–P 9, 2C, CH3–CH2–O), 70.1
(Cl3C).
N-tert-Butyl-N-[2-ethyl-(1-diethoxyphosphoryl)hexyl]-
O-(1-phenylethyl) Hydroxylamine 6
PMDETA (1.6 mL, 9.0 mmol) was added to a degassed solu-
tion of CuBr (646 mg, 4.5 mmol) and copper (285 mg, 4.5 mmol)
in dichloromethane, and nitrogen was bubbled through the
solution for 10 min. A degassed dichloromethane solution of
ETHEXNO 3d (1.0 g, 3.0 mmol) and 1-bromoethylbenzene
(441 mg, 4.5 mmol) was transferred to the mixture, which was
then stirred for 2 h at room temp under nitrogen. Diethyl ether
(30 mL) was added and the solid filtered off. The organic
layer was washed with water until colourless and then dried
with MgSO4. The solvent was removed to yield a colourless
oil (920 mg, 70%). δP 26.17 (33.2%), 26.66 (14.1%), 26.80
(17.5%), 27.41 (35.2%). δH 0.73–0.97 (m, 6H, CH3CH2), 1.18
(s, 9H, CCH3), 1.24–1.44 (m + t, 12H, JH–H 7.1, 2 × CH3–
CH2–O + 3 × CH2), 1.54 (d, 3H, 1 × CH*–CH3), 1.74–2.16 (m,
3H, 1 × CH2 + 1 × CH), 3.35–3.51 (dd, 1H, JH–H 10.4, CH),
3.97–4.39 (m, 4H, 2 × CH3–CH2–O), 5.08–5.22 (m, 1H, CH*–
CH3), 7.15–7.43 (m, 5H, –CH=CH–). δC 12.8 (CH3CH2),
15.7 (CH3CH2CH2), 20.8 (CH3CH2CH2), 22.6 (CH2CH2CH2),
24.1 (CH3CH2CH2), 25.7 (CH3CH2CH), 26.4 (3C, C(CH3)3),
30.1 (CH2CH2CH2), 48.3 (CH3–CH2–O), 50.2 (CHCH2), 55.2
(CHCH), 59.8 (C(CH3)3), 62.5 (CH3–CH2–O), 79.4 (CH–NH),
82.9 (CHCH3), 125.1–12.6 (5C, CH=CH), 145.2 (C).
Diethyl [1-(tert-Butylamino)-2-ethylhexy] Phosphonate 2d
Imine 1d (0.054 mol) and diethyl phosphate (0.070 mol) were
cooled at 0◦C. BF3·OEt (0.005 mol) was then added. The reac-
tion mixture was stirred for an hour at room temperature.
It was then filtered, dissolved in ether, and the residue was
washed two times with 150 mL of 5% HCl in water. The
solution was extracted three times with ether. Na2CO3 was
added to the combined organic layers to give a basic solution
(pH = 8). The solution was then extracted three times with ether.
The organic phase was dried over MgSO4. The solvent was
removed under reduced pressure to give a colourless oil (16.5 g,
94%). δH 0.85–1.00 (m, 6H, CH3CH2), 1.15 (s, 9H, C–CH3),
1.28–1.40 (t + m, 12H, JH–H 7.1, 2 × CH3–CH2–O + 3 × CH2),
1.5–1.65 (m, 3H, 1 × CH2 + 1 × CH), 3.15 (d, 1H, JH–H
19.6, CH), 4.05–4.25 (m, 4H, 2 × CH3–CH2–O). δC 12.6
(CH3CH2), 13.0 (CH3CH2CH2), 22.8–23.1 (CH3CH2CH2),
23.2–23.3 (CH2CH2CH2), 29.8–29.9 (CH2CH2CH2), 30.5–
30.6 (CH3CH2CH), 30.3 (3 × CH3C), 32.6 (CH2CH2CH2),
2-Methyl-2-[N-tert-butyl-N-(2-ethyl-(1-
diethoxyphosphoryl)hexyl)]propanoic Acid 7
PMDETA (1.6 mL, 9.0 mmol) was added to a degassed
solution of CuBr (646 mg, 4.5 mmol) and copper (285 mg,
4.5 mmol) in dichloromethane, and nitrogen was bubbled

RESEARCH FRONT
1244
J. Marchand et al.
through the solution for 10 min. A degassed dichloromethane
solution of ETHEXNO 3d (1.0 g, 3.0 mmol) and 2-bromo-2-
methylpropionic acid (745 mg, 4.5 mmol) was transferred to the
mixture which was then stirred for 1 h at room temperature under
nitrogen. Diethyl ether (30 mL) was added and the solid fil-
tered off. The organic layer was washed with 5% HCl in water
two times and with water until colourless and then dried with
MgSO4. The solvent was removed to yield an oil which precip-
itated in cold pentane (755 mg, 60%). δP 29.50 (60%), 29.22
(13%), 26.65 (27%). δH 0.76–1.02 (m, 8H, CH3CH2 + CH2),
1.18 (s, 9H, CCH3), 1.27–1.37 (m, 12H, 2 × CH3–CH2–
O + 3 × CH2), 1.51–1.56 (d, 6H, JH–H 8.8, 2 × CH3–CH2–
O), 1.73 (s, 1H, CH–P), 3.35–3.52 (m, 1H, JH–H 10.5,
CH–CH), 4.05–4.25 (m, 4H, 2 × CH3–CH2–O). δC 9.9 (1C,
CH3CH2), 13.1 (1C, CH3CH2CH2), 15.5 (1C, CH3CH2CH2),
22.1 (1C, CH2CH2CH2), 23.8 (1C, CH2CH2CH2), 25.6 (1C,
CH3CH2CH), 27.6 (2C, HOOCC(CH3)2), 29.0 (3 × C(CH3)3),
33.1 (2C, CH3–CH2–O), 49.3 (1C, CHCH2), 59.7 (1C, CH–
NO), 63.8 (2C, CH3–CH2–O), 82.7 (C(CH3)3), 91.2 (1C,
CCOOH), 175.6 (1C, C=O). m/z (ESI) 423 (M+).
[17] F. Chauvin, J.-L. Couturier, P.-E. Dufils, P. Gérard, D. Gigmes,
O. Guerret, Y. Guillaneuf, S. R. A. Marque, D. Bertin, P. Tordo, ACS
Symp. Ser. 2006, 944, 326. doi:10.1021/BK-2006-0944.CH023
[18] P. Gerard, L. Couvreur, S. Magnet, J. Ness, S. Schmidt, ACS Symp.
Ser. 2009, 1024, 361. doi:10.1021/BK-2009-1024.CH024
[19] D. Gigmes, S. Marque, O. Guerret, J.-L. Couturier, F. Chauvin,
P.-E. Dufils, D. Bertin, P. Tordo, WO Patent 2004/014926 2004.
[20] Y. Guillaneuf, J.-L. Couturier, D. Gigmes, S. R. A. Marque, P. Tordo,
D. Bertin, J. Org. Chem. 2008, 73, 4728. doi:10.1021/JO800422A
[21] P. E. Dufils, N. Chagneux, D. Gigmes, T. Trimaille, S. R. A.
Marque, D. Bertin, P. Tordo, Polymer 2007, 48, 5219. doi:10.1016/
J.POLYMER.2007.06.050
[22] Y. Guillaneuf, P.-E. Dufils, L. Autissier, M. Rollet, D. Gigmes,
D. Bertin, Macromolecules 2010, 43, 91. doi:10.1021/MA901838M
[23] D. Gigmes, J.Vinas, N. Chagneux, C. Lefay,T. N.T. Phan,T.Trimaille,
P.-E. Dufils, Y. Guillaneuf, G. Carrot, F. Boué, D. Bertin, ACS Symp.
Ser. 2009, 1024, 245. doi:10.1021/BK-2009-1024.CH016
[24] D. Gigmes, D. Bertin, C. Lefay,Y. Guillaneuf, Macromol.Theor. Simul.
2009, 18, 402. doi:10.1002/MATS.200900019
[25] H. Fischer, Chem. Rev. 2001, 101, 3581. doi:10.1021/CR990124Y
[26] S. Marque, J. Org. Chem. 2003, 68, 7582. doi:10.1021/JO030036K
[27] S. Marque, J. Sobek, H. Fischer, A. Kramer, P. Nesvadba,
W. Wunderlich, Macromolecules 2003, 36, 3440. doi:10.1021/
MA021605D
[28] S. Acerbis, E. Beaudouin, D. Bertin, D. Gigmes, S. Marque,
P. Tordo, Macromol. Chem. Phys. 2004, 205, 973. doi:10.1002/
MACP.200300240
Acknowledgements
The authors acknowledge ARKEMA, the University of Provence and the
CNRS for financial support.
[29] D. Gigmes, S. Marque, D. Bertin, Recent Res. Dev. Org. Chem. 2006,
10, 63.
References
[1] W. A. Braunecker, K. Matyjaszewski, Prog. Polym. Sci. 2007, 32, 93.
doi:10.1016/J.PROGPOLYMSCI.2006.11.002
[30] J. Sobek, R. Martschke, H. Fischer, J. Am. Chem. Soc. 2001, 123,
2849. doi:10.1021/JA0036460
[2] M. K. Georges, R. P. N. Veregin, P. M. Kazmaier, G. K. Hamer,
Macromolecules 1993, 26, 2987. doi:10.1021/MA00063A054
[3] C. J. Hawker, A. W. Bosman, E. Harth, Chem. Rev. 2001, 101, 3661.
doi:10.1021/CR990119U
[31] Y. Guillaneuf, D. Bertin, P. Castignolles, B. Charleux, Macromolecules
2005, 38, 4638. doi:10.1021/MA050241X
[32] K. O. Siegenthaler, A. Studer, Macromolecules 2006, 39, 1347.
doi:10.1021/MA0513463
[4] D. H. Solomon, E. Rizzardo, P. Cacioli, Polymerization process and
polymers produced thereby. US Patent 4581429 1986 [Chem. Abs.
1985, 102, 221335q].
[5] V. Sciannamea, R. Jerome, C. Detrembleur, Chem. Rev. 2008, 108,
1104. doi:10.1021/CR0680540
[33] H. Fischer, S. R. A. Marque, P. Nesvadba, Helv. Chim. Acta 2006, 89,
2330. doi:10.1002/HLCA.200690217
[34] D. Zubenko, Y. Tsentalovich, N. Lebedeva, I. Kirilyuk, G. Roshchup-
kina, I. Zhurko,V. Reznikov, S. R.A. Marque, E. Bagryanskaya, J. Org.
Chem. 2006, 71, 6044. doi:10.1021/JO060787X
[6] M. Kamigaito, T. Ando, M. Sawamoto, Chem. Rev. 2001, 101, 3689.
doi:10.1021/CR9901182
[35] H. Fischer, ACS Symp. Ser. 2003, 854, 10. doi:10.1021/BK-2003-
0854.CH002
[7] K. Matyjaszewski, J. H. Xia, Chem. Rev. 2001, 101, 2921. doi:10.1021/
CR940534G
[8] J. S. Wang, K. Matyjaszewski, J. Am. Chem. Soc. 1995, 117, 5614.
doi:10.1021/JA00125A035
[9] M. Kato, M. Kamigaito, M. Sawamoto, T. Higashimura, Macro-
molecules 1995, 28, 1721. doi:10.1021/MA00109A056
[10] J. Chiefari, Y. K. Chong, F. Ercole, J. Krstina, T. P. T. Le, R. T. A.
Mayadunne, G. F. Meijs, G. Moad, E. Rizzardo, S. H. Thang,
Macromolecules 1998, 31, 5559. doi:10.1021/MA9804951
[11] G. Moad, E. Rizzardo, S. H. Thang, Aust. J. Chem. 2005, 58, 379.
doi:10.1071/CH05072
[36] H. Fischer, M. Souaille, Chimia 2001, 55, 109.
[37] C. Le Mercier, S. Acerbis, D. Bertin, F. Chauvin, D. Gigmes,
O. Guerret, M. Lansalot, S. Marque, F. Le Moigne, H. Fischer,
P. Tordo, Macromol. Symp. 2002, 182, 225. doi:10.1002/1521-3900
(200206)182:1<225::AID-MASY225>3.0.CO;2-X
[38] F. Chauvin, P. E. Dufils, D. Gigmes, Y. Guillaneuf, S. R. A. Marque,
P. Tordo, D. Bertin, Macromolecules 2006, 39, 5238. doi:10.1021/
MA0527193
[39] K. Matyjaszewski, B. E. Woodworth, X. Zhang, S. G. Gaynor,
Z. Metzner, Macromolecules 1998, 31, 5955. doi:10.1021/
MA9807264
[12] A. Favier, M.T. Charreyre, Macromol. Rapid Commun. 2006, 27, 653.
doi:10.1002/MARC.200500839
[13] S. Perrier, P. Takolpuckdee, J. Polym. Sci., PartA: Polym. Chem. 2005,
43, 5347. doi:10.1002/POLA.20986
[14] D. Benoit, V. Chaplinski, R. Braslau, C. J. Hawker, J. Am. Chem. Soc.
1999, 121, 3904. doi:10.1021/JA984013C
[15] D. Benoit, G. Grimaldi, S. Robin, J.-P. Finet, P. Tordo, Y. Gnanou,
J. Am. Chem. Soc. 2000, 122, 5929. doi:10.1021/JA991735A
[16] S. Grimaldi, J.-P. Finet, F. Le Moigne, A. Zeghdaoui, P. Tordo,
D. Benoit, M. Fontanille,Y. Gnanou, Macromolecules 2000, 33, 1141.
doi:10.1021/MA9913414
[40] D. Bertin, D. Gigmes, S. Marque, P. Tordo, E-Polymers 2003, 002.
[41] S. Marque, C. Le Mercier, P.Tordo, H. Fischer, Macromolecules 2000,
33, 4403. doi:10.1021/MA9918452
[42] E. Drockenmuller, J.-P. Lamps, J.-M. Catala, Macromolecules 2004,
37, 2076. doi:10.1021/MA0351221
[43] S. Marque, H. Fischer, E. Baier, A. Studer, J. Org. Chem. 2001, 66,
1146. doi:10.1021/JO001190Z