Unique behavior of nitroxide biradicals in the controlled radical polymerization of styrene

Article abstract of DOI:10.1021/ma0116856

The use of various binitroxides, with both radical sites of similar reactivity, as mediator in the controlled radical polymerization of styrene was examined and the rate constants of the reactions involved were determined at 130 °C. Typical features of a controlled radical polymerization were observed in the early stage of the polymerization, leading to the formation of two-arm macromolecules containing the binitroxide at the core. At higher conversion, however, continuous decomposition reaction resulted in a break of the two-arm macromolecules into a dead chain having an unsaturation end group and a living chain capped by a modified binitroxide. The latter had the free nitroxide site inactivated by conversion into hydroxylamine. The extent of this side reaction was much larger than in classical nitroxide-mediated controlled radical polymerization of styrene. This feature was assigned to the structure of the growing chains. As they contain the binitroxide at the core, the activation reaction produces a propagating chain and a nitroxide to which another polymeric chain remains attached by the second alkoxyamine bond. Thus, deactivation of the propagating radical by nitroxide is a bimolecular process between two macromolecular species. This unique situation had no significant effect on the rate constant of the alkoxyamine homolytic dissociation, but more importantly decreased the rate constant of recombination. The latter was 30 times lower than that determined for the recombination of TEMPO with polystyryl radical at 130 °C. The slow recombination was assigned to steric hindrance and resulted in an unusually high concentration of nitroxide in the polymerization medium, together with a large concentration of propagating radicals. A consequence of the high concentration of both radicals was the enhanced rate of hydrogen transfer from the active species to the persistent radical, leading to alkoxyamine bond breaking and, hence, to arm separation.

Full text of DOI:10.1021/ma0116856

Macromolecules 2002, 35, 2305-2317
2305
Unique Behavior of Nitroxide Biradicals in the Controlled Radical
Polymerization of Styrene
Wen li Hu a n g,† Rober t Ch ia r elli,† Ber n a d ette Ch a r leu x,*,‡ An d r e´ Ra ssa t,† a n d
J ea n -P ier r e Va ir on ‡
De´partement de Chimie, Unite´ Mixte associe´e au CNRS, UMR8640, Ecole Normale Supe´rieure,
24, rue Lhomond, 75231 Paris Cedex 05, France; and Laboratoire de Chimie Macromole´culaire,
Unite´ Mixte associe´e au CNRS, UMR 7610 Universite´ Pierre et Marie Curie, Tour 44,
1er e´tage, 4, Place J ussieu, 75252 Paris Cedex 05, France
Received September 27, 2001
ABSTRACT: The use of various binitroxides, with both radical sites of similar reactivity, as mediator in
the controlled radical polymerization of styrene was examined and the rate constants of the reactions
involved were determined at 130 °C. Typical features of a controlled radical polymerization were observed
in the early stage of the polymerization, leading to the formation of two-arm macromolecules containing
the binitroxide at the core. At higher conversion, however, continuous decomposition reaction resulted in
a break of the two-arm macromolecules into a dead chain having an unsaturation end group and a living
chain capped by a modified binitroxide. The latter had the free nitroxide site inactivated by conversion
into hydroxylamine. The extent of this side reaction was much larger than in classical nitroxide-mediated
controlled radical polymerization of styrene. This feature was assigned to the structure of the growing
chains. As they contain the binitroxide at the core, the activation reaction produces a propagating chain
and a nitroxide to which another polymeric chain remains attached by the second alkoxyamine bond.
Thus, deactivation of the propagating radical by nitroxide is a bimolecular process between two
macromolecular species. This unique situation had no significant effect on the rate constant of the
alkoxyamine homolytic dissociation, but more importantly decreased the rate constant of recombination.
The latter was 30 times lower than that determined for the recombination of TEMPO with polystyryl
radical at 130 °C. The slow recombination was assigned to steric hindrance and resulted in an unusually
high concentration of nitroxide in the polymerization medium, together with a large concentration of
propagating radicals. A consequence of the high concentration of both radicals was the enhanced rate of
hydrogen transfer from the active species to the persistent radical, leading to alkoxyamine bond breaking
and, hence, to arm separation.
Sch em e 1. (A) Alk oxya m in e Mu ltifu n ction a l In itia tor
(R ) Alk yl Gr ou p ; X ) In itia tor F r a gm en t;
D ) P olym er Ch a in ) a n d (B) Alk oxya m in e Ba sed
on a Mu ltifu n ction a l Nitr oxid e (Mu ltifu n ction a l
Cou p lin g Agen t)
In tr od u ction
Nitroxide-mediated controlled free-radical polymeri-
zation (CRP) is based on an equilibrium between
propagating macromolecular radicals and dormant chains
that have an alkoxyamine end functionality.1-4 Provid-
ing the nitroxide is carefully selected so as to allow the
homolytic cleavage of the NO-C bond under acceptable
conditions, polymerization yields well-defined polymers
with predictable molar mass and narrow molar mass
distribution. The chains formed exhibit an end func-
tionality close to 1 and are able to further grow upon
new addition of monomer at the polymerization tem-
perature. Many types of complex architectures have
been created, as an additional proof of the “livingness”
of the system.1,5 For this goal, preformed alkoxyamines
were used as well-defined mono, bi or multifunctional
initiators, leading respectively to single-arm, biarm, or
multiarm structures (Scheme 1A). The most studied
nitroxide, TEMPO (2,2,6,6-tetramethylpiperidine-N-
oxyl), provides good control for the polymerization of
styrenic monomers at elevated temperatures (T > 110
°C).1 Other cyclic nitroxides have been examined, not
behaving differently from TEMPO.1 A new generation
of acyclic nitroxides bearing a hydrogen on the carbon
R to the nitroxide function are more efficient at control-
ling a broader range of monomers at significantly lower
temperatures and with faster kinetics.6,7 Among all the
studies in the field, the use of nitroxide biradicals or
multiradicals (as previously, we distinguish multiradi-
cals, small molecules containing several nitroxide groups,
from polyradicals, polymers made of monoradical mono-
* To whom correspondence should be addressed. E-mail:
charleux@ccr.jussieu.fr.
† Ecole Normale Supe´rieure.
‡ Universite´ Pierre et Marie Curie.
10.1021/ma0116856 CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/13/2002

2306 Huang et al.
Macromolecules, Vol. 35, No. 6, 2002
mers)8 as control agents in radical polymerization has
never been reported. Such species, and more particu-
larly binitroxides, have long been known9 and used as
spin probes10 and more recently as magnetic material11
or spin-labeled spin traps.12 In addition to their ability
to control radical polymerizations, nitroxide multiradi-
cals with sites of similar reactivity can be seen as
reversible multifunctional coupling agents. Compared
with the usual alkoxyamine multifunctional initiators,
these nitroxides and the derived alkoxyamines might
offer different synthetic capabilities (Scheme 1). The
arms would grow as single chains, independent of the
structure and size of the overall macromolecule, pre-
venting star-star coupling via propagating radical ter-
minations. They would exchange from one macromol-
ecule to another, ensuring a narrow molar mass
distribution of the final polymer. The resulting multi-
arm macromolecules would not be thermally stable and
therefore able to release the arms under selected condi-
tions. For instance, methacrylic- and styrenic-based
monomers with a nitroxide function were used as both
comonomers and mediators in the CRP of styrene.13 The
resulting polymeric nitroxide led to branched structures,
whose formation is thermally reversible. In contrast, if
a nitroxide with very low reactivity to carbon radical
existed, combining this inert group with a reactive
nitroxide in the same molecule would permit monitoring
of the chain growth and end group purity by ESR12
during the controlled polymerization.
Sch em e 2. Nitr oxid e Bir a d ica ls B1-B5 a n d
Bisa lk oxya m in e A1
Service of Microanalysis, University Pierre and Marie Curie
(Paris). Nuclear magnetic resonance (NMR) spectra were
recorded on Bruker AC spectrometers operating at 200 or 250
1
MHz for H and 62.9 MHz for 13C, using deuterated chloroform
as the solvent and the internal solvent peak as the reference.
Size exclusion chromatography (SEC) was carried out using a
Waters apparatus working at room temperature with stabi-
lized tetrahydrofuran eluent at a flow rate of 1 mL‚min-1 and
equipped with four PL-gel 10µ columns (100, 500, 103, and 104
Å). A differential refractive index detector was used and molar
The purpose of this work was to study the behavior
of various binitroxides with sites of similar reactivity
in the radical polymerization of styrene, and to examine
their ability to act as both coupling and control agents.
masses were derived from
a calibration curve based on
polystyrene standards. The concentration of chains in the
polymerization medium was calculated according to the fol-
lowing formula
Exp er im en ta l Section
Ma ter ia ls. Styrene (S; 99% from Aldrich) was distilled
under reduced pressure before use. Benzoyl peroxide (BPO;
97% from Aldrich) was recrystallized at room temperature
from a chloroform/methanol mixture. TEMPO (98% from
Aldrich) was purified by sublimation. m-Chloroperbenzoic acid
(mCPBA; 77% from Aldrich, remainder 3-chlorobenzoic acid
and water) was purified as described.14 m-Xylene (Aldrich) and
tert-butylbenzene (Aldrich) were stirred three times with 15
vol % of concentrated H2SO4 for 30 min; then they were washed
twice with H2O, then with 10% aqueous Na2CO3 and finally
with H2O; after drying over Na2SO4, they were distilled, and
stored over molecular sieves (3 Å).15 Tetrahydrofuran (THF)
was distilled first from potassium hydroxide and then from
Na-benzophenone immediately before use. Unless otherwise
specified, other materials were purchased from commercial
suppliers and used without further purification. Organic
solutions were dried over Na2SO4 and evaporated using a
rotary evaporator. For synthesis of the binitroxides, the
sentence “the oxidation was monitored by ESR” means that,
after each addition of oxidant, a sample was taken from the
reaction mixture and its ESR spectrum was recorded; when
the intensity of the radical signal (measured by double
integration) reached a plateau, the addition of mCPBA was
discontinued.
xdS
[chains] (mol‚L-1) )
(I)
Mn
with dS ) the styrene density (909 g‚L-1), x ) the conversion
of monomer and Mn ) the number average molar mass
(g‚mol-1).
Electron spin resonance (ESR) spectra were recorded on a
Bruker ESR300 spectrometer operated in continuous wave
(CV) mode with a TE102 rectangular cavity. The modulation
amplitude was kept constant at 1 G, while the microwave
power was held at 10 mW. For the ESR studies at 130 °C, the
tube was heated in an oil bath and then cooled by cold water
and analyzed at room temperature in the cavity of the
spectrometer. The relative concentration of nitroxide radicals
was measured by double integration of the ESR spectrum. The
number of spin s in the solution (absolute radical concentra-
tion) was calculated by comparing the spectrum of a solution
of sublimed TEMPO of known concentration, recorded under
strictly the same experimental conditions. Under these condi-
tions, traces of residual biradical, although not visually
apparent in the spectrum,12 would contribute to this number
of spin.
An a lytica l Tech n iqu es. Thin-layer chromatography (TLC)
was performed on commercial Merck plates coated with SiO2
or neutral Al2O3 (60F254; 0.25 mm thick). Preparative column
chromatography was carried out using SiO2 60 (0.063-0.2 mm)
or activated neutral Al2O3 III (from Merck). For some samples,
preparative chromatography was performed on plates coated
Syn th esis of th e Nitr oxid e Bir a d ica ls a n d of th e
Bisa lk oxya m in e (Sch em e 2). 1,9-Dioxy-2,2,8,8,10,10-h ex-
a m eth yl-1,9-d ia za sp ir o[5.5]u n d eca n e-4-on e (B1) was pre-
pared as described, from the diaminoketone 2,2,8,8,10,10-
hexamethyl-1,9-diazaspiro[5.5]undecane-4-one (1).16 Rf ) 0.75
(Al2O3, Et2O); MS (EI) m/z 282 (M), 267 (M - CH3), 237 (M -
3CH3), 222 (M - 4CH3); mp 112.5-114 °C (lit.:16 111-113 °C).
Anal. Calcd for C15H26N2O3 (282.4): C, 63.80; H, 9.28; N, 9.92.
Found: C, 63.77; H, 9.30; N, 9.80. ESR: a single broad line
(27 G) at room temperature in Et2O (M/1000).
with SiO2 (60F254; 2 mm thick) or with neutral Al2O3 (60F254
;
1.5 mm thick). EI (electrical impact) and CI (chemical impact,
ammonia) mass spectra were obtained with a Nermag R10-
10C spectrometer or a J EOL MS 700 spectrometer. Melting
points were measured on a Bu¨chi 535 melting point apparatus.
Elemental analyses (C, H, and N) were performed at the
1,9-Dioxy 2,2,8,8,10, 10-h exa m eth yl-1,9-d ia za sp ir o[5.5]-
u n d eca n -4-ol (B2). A solution of the diaminoketone 1 (1.54

Macromolecules, Vol. 35, No. 6, 2002
Unique Behavior of Nitroxide Biradicals 2307
g, 6.1 mmol) in anhydrous THF (75 mL) was added dropwise
to a solution of LiAlH4 (465 mg, 12.2 mmol) in anhydrous THF
(110 mL). After the mixture was stirred at room temperature
for 4 h, H2O (0.45 mL), 15% aqueous NaOH (0.45 mL), and
H2O (1.35 mL) were successively added dropwise while stir-
ring. The resulting suspension was filtered through Celite,
rinsed with a mixture of CH2Cl2/MeOH (99/1) and the filtrate
was dried and evaporated to give the diaminoalcohol 2
(2,2,8,8,10,10-hexamethyl-1,9-diazaspiro[5.5]undecan-4-ol) (1.45
g, 5.7 mmol) as a yellow solid (93%). 13C NMR (62.9 MHz): δ
28.5, 30.6, 34.1, 34.3, 34.8 (CH3); 45.7, 48.2, 51.1, 53.3 (CH2);
50.1, 50.3, 51.5, 54.0 (C); 63.1 (CH). A solution of mCPBA (3.01
g, 17.6 mmol) in CH2Cl2 (50 mL) in 5 mL fractions was added
to a solution of 2 (1.45 g, 5.7 mmol) in CH2Cl2 (300 mL). After
complete oxidation (monitored by ESR), the solution was
treated with aqueous NaHCO3 (5%); the organic phase was
separated, dried, and evaporated. The crude product was
purified by column chromatography on Al2O3, using first Et2O
then Et2O/MeOH (98/2) as an eluent, to give B2 (810 mg, 2.9
mmol) as an orange-red solid (50%). Rf ) 0.53 (Et2O/MeOH,
98/2); MS (EI) m/z 284 (M), 269 (M - CH3), 239 (M - 3 CH3),
224 (M - 4 CH3); mp 110.0-111.5 °C. Anal. Calcd for
0.9 mmol) as a white solid (89%). Rf ) 0.45; 1H NMR (200 MHz)
δ 0.2-3 (m, 8 CH3 and 4 CH2; 32 H), 3.83 and 4.6 (m, 2 CH;
2 H), 7.0-7.5 (m, aromatic protons; 10 H). Anal. Calcd for
C
31H44N2O3 (492.70): C, 75.57; H, 9.00; N, 5.68. Found: C,
75.08; H, 9.29; N, 5.41. MS (CI): m/z 495 (M + 2), 496 (M +
3), 391 (M + 3 - C8H8), 375 (M + 3 - C8H8 - O).
2,2,6,6-Tetr a m eth ylp ip er id in e-N-(1-p h en eth yloxy) (S-
TEMP O) was prepared in the presence of J acobsen’s reagent
according to literature.18 Complete characteristics of this
molecule have already been reported.18,19
Bu lk P olym er iza tion of Styr en e. The solutions of BPO
(73 mg; 0.030 mol‚L-1) and binitroxide (0.018 mol‚L-1, [nitrox-
ide group, NO•] ) 0.036 mol‚L-1) in styrene (9.1 g; 10 mL) were
degassed by three freeze-pump-thaw cycles. The tubes were
sealed off under vacuum and then immersed in an oil bath
thermostated at 110 or 130 °C. After polymerization, the
content of each tube was poured while stirring into an excess
of methanol (200 mL). The polymers were isolated by filtration,
washed with methanol and dried. Styrene conversion (x) was
determined by gravimetry. The same procedure was applied
for alkoxyamine initiated polymerizations. Results are re-
ported in Tables 1 and 2.
The methanol filtrates and the purified polymers (dissolved
in a solution of THF or CH2Cl2) were analyzed by ESR. In all
cases, a spectrum characteristic of that of a mononitroxide was
observed (see below). The concentration of nitroxide present
in the filtrates was always very small because of dilution and
was not used for further calculation. The number of spin S
(mol‚L-1) of nitroxide attached to a polymer chain in the
polymerization reaction medium was calculated from the
number of spin s of a solution of 10 mg of polymer dissolved
in 2 mL of THF (or CH2Cl2), according to the following
relationship
C
15H28N2O3 (284.4): C, 63.35; H, 9.92; N, 9.85. Found: C,
63.31; H, 10.11; N, 9.70. ESR: a single broad line (27 G) at
room temperature in Et2O (M/1000).
1,9-Dioxy-2,2,8,8,10,10-h exa m eth yl-1,9-d ia za sp ir o[5.5]-
u n d eca n e (B3). Hydrazine hydrate (1 mL) was added to a
solution of diaminoketone 1 (1.5 g, 6.0 mmol) in diethylene
glycol (5 mL) in a flask equipped with a reflux condenser with
variable takeoff. The mixture was refluxed for 2 h. After
cooling to room temperature, KOH pellets (1.96 g) were added
and H2O distilled off. When the distillation temperature rose
above 100 °C, the takeoff was disconnected and the mixture
was refluxed for 1.5 h. The mixture was allowed to cool to room
temperature, H2O (5 mL) was added and the reflux condenser
replaced by a distillation bridge. Water was distilled off, and
the distillate was extracted with Et2O (3 × 5 mL). The organic
phases were combined, dried and evaporated to give crude
diamine 3 (2,2,8,8,10,10-hexamethyl-1,9-diazaspiro[5.5]unde-
cane) (460 mg) as a bright yellow liquid. This product was
oxidized directly without further purification. A solution of
mCPBA (250 mg, 1.5 mmol) in CH2Cl2 (5 mL) in 0.5 mL
fractions was added to a solution of crude diamine 3 (107 mg,
0.5 mmol) in CH2Cl2 (20 mL). After complete oxidation
(monitored by ESR), the solution was treated with aqueous
NaHCO3 (5%). The organic phase was separated, dried, and
evaporated. The crude product was purified by chromatogra-
phy with a SiO2 plate, using Et2O/pentane (50/50) as an eluent
to give the biradical B3 (50 mg, 0.18 mmol) as an orange-red
solid (13% yield for two steps). Rf ) 0.50; MS (EI) m/z 268
(M), 253 (M - CH3), 237 (M - 2 CH3 - 1); mp 43-44 °C. Anal.
Calcd for C15H28N2O2 (268.4): C, 67.13; H, 10.52; N, 10.44.
Found: C, 67.27; H, 10.56; N, 10.24; ESR: a single broad line
(27 G) at room temperature in Et2O (M/1000).
2
10
WPS‚s‚
2
S )
) s‚xd
(II)
WS
dS
S10
with WPS and WS, the respective weight of polystyrene and
initial weight of styrene in the polymerization medium. The
ESR results are reported in Table 3.
Dir ect ESR Mon itor in g of th e Nitr oxid e Ra d ica ls
F or m ed d u r in g th e Cou r se of P olym er iza tion . As de-
scribed above, the quantity of nitroxide radicals attached to a
polymer chain and formed during polymerization can be
estimated from a solution of the precipitated polymer, obtained
at different polymerization times. To get an accurate overall
value (including the polymeric and the small nitroxides) and
avoid side reactions (such as spontaneous transformation of
hydroxylamine into nitroxide in air), a direct ESR analysis was
made on the reaction mixture. A known concentration of a
solution of bisalkoxyamine A1 in styrene (1.0 × 10-2 and 2.0
× 10-2 mol‚L-1 for the two respective experiments) was
introduced in an ESR tube. The tube was thoroughly degassed
by three freeze-pump-thaw cycles and sealed off under
vacuum. It was then heated at 130 °C in a thermostated oil
bath. At regular time intervals, the tube was removed from
the bath and the concentration of nitroxide radicals produced
during the course of the reaction was monitored by ESR at
room temperature. Figure 1 shows the spectra obtained after
heating during (a) 15 min, (b) 2.5 h, (c) 5 h, and (d) 10 h. At
the beginning of the experiment (a and b), the three-line
spectra characteristic of a mononitroxide in solution were
observed with a rotational correlation time20 τc ) 0.5 × 10-10
s after 15 min, increasing to τc ) 5 × 10-10 s after 2.5 h. After
5 h (c), a broad shoulder is detected on the low-field line,
indicating the presence of two species: the three-line spectrum
corresponds to a mononitroxide with τc ) 5 × 10-10 s, while
the broad shoulder may be attributed by comparison with
reference spectra, to a mononitroxide with a τc of the order of
5-7 × 10-9 s. After 10 h (d), a new signal appeared at the
lower field. Again two species are present, one with τc ) 5 ×
10-10 s and the other one with τc of the order of 10-8 s. No
Hexa n ed ioic a cid bis(1-oxy-2,2,6,6-tetr a m eth ylp ip er i-
d in -4-yl) ester (B4) and d eca n ed ioic a cid bis(1-oxy-
2,2,6,6-tetr a m eth ylp ip er id in -4-yl) ester (B5) were pre-
pared as described.17
2,2,8,8,10,10-Hexa m eth yl-1,9-bis(1-p h en yleth oxy)-1,9-
diazaspir o[5.5]u n decan -4-on e (A1). The synthesis of alkoxy-
amines in the presence of J acobsen’s reagent was described
by Dao et al.18 and this method was used here to prepare A1.
[N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediami-
no]manganese(III) chloride (J acobsen’s catalyst) (280 mg, 0.4
mmol) was added to a solution of biradical B1 (284 mg, 1.0
mmol) and styrene (419 mg, 4.0 mmol) in 1/1 toluene/ethanol
(15 mL). This step was followed by the addition of di-tert-butyl
peroxide (0.54 mL, 3.0 mmol) and sodium borohydride (228
mg, 6.0 mmol). The reaction mixture was then stirred at room
temperature for 4 h, evaporated to dryness, and partitioned
between CH2Cl2 (15 mL) and H2O (20 mL). The aqueous layer
was further extracted with CH2Cl2 (3 × 20 mL). The organic
layers were combined, dried, and evaporated. The crude
product was purified by column chromatography on Al2O3
(pentane/ether, 50/50) to give the bisalkoxyamine A1 (444 mg,

2308 Huang et al.
Macromolecules, Vol. 35, No. 6, 2002
Ta ble 1. Bu lk P olym er iza tion of Styr en e a t 130 a n d 110 °C (/), In itia ted by BP O (0.030 m ol‚L-1) in th e P r esen ce of
Va r iou s Nitr oxid e Bir a d ica ls ([NO•] ) 0.036 m ol‚L-1): SEC An a lysis of th e P olym er s
raw polymera
heated polymerb
experimental
theoretical
experimentalc
time
(h)
Mn,th
Mn,th/2
Mn
Mp1
Mp2
[chain]
Mn
binitroxide
convn
(g‚mol-1
)
(g‚mol-1
)
(g‚mol-1
)
(g‚mol-1
)
(g‚mol-1
)
Mw/Mn
(mol‚L-1
)
(g‚mol-1
)
Mw/Mn
B1
2.5
5.0
14.0
2.5
5.0
14.0
2.5
5.0
14.0
2.5
5.0
14.0
2.5
0.38
0.60
0.89
0.12
0.29
0.74
0.22
0.61
0.84
0.43
0.64
0.85
0.53
0.72
0.88
0.53
0.76
0.86
19000
30200
44800
6060
9500
15100
22400
3030
7320
18750
5650
15400
21160
10950
16240
21340
13400
18200
22250
13250
19200
21700
17300
23900
31200
7400
27900
46600
1.34
1.44
1.48
1.31
1.40
1.56
1.29
1.43
1.42
1.41
1.55
1.62
1.37
1.44
1.57
1.68
1.60
1.72
0.020
0.023
0.026
0.015
0.019
0.021
0.019
0.023
0.029
0.021
0.024
0.031
0.022
0.026
0.031
0.023
0.026
0.031
11100
13900
21400
3840
9000
21600
6500
14500
21000
12500
19300
19800
1.25
1.37
1.69
1.55
1.41
1.45
1.22
1.29
1.51
1.32
1.32
1.51
26500
45200
B1*
B2
B3
B4
B5
9500
22000
57000
15700
46200
73000
32000
55200
75300
14640
37500
11300
30900
42300
21900
32470
42670
26800
36400
44500
26500
38400
43400
14200
32000
10500
24000
26000
19000
24300
24900
22200
25300
26000
20600
26300
25000
23400
35100
16100
25800
34200
5.0
14.0
2.5
5.0
14.0
a
b
c
Polymer recovered after precipitation in methanol, washed and dried under vacuum. Polymer heated at 130 °C for 24 h. Mpi:
mass at the maximum of the ith peak.
Ta ble 2. Bu lk P olym er iza tion of Styr en e a t 130 °C, in th e
P r esen ce of th e Bisa lk oxya m in e A1: SEC An a lysis of
P olym er s
Ta ble 3. Bu lk P olym er iza tion of Styr en e a t 130 a n d 110
°C (/), In itia ted by BP O (0.030 m ol‚L-1) in th e P r esen ce
of Va r iou s Nitr oxid e Bir a d ica ls ([NO•] ) 0.036 m ol‚L-1):
ESR An a lysis of th e Ra w P olym er befor e a n d a fter
Oxid a tion w ith P bO2 a t Room Tem p er a tu r e in CH2Cl2
theoretical
Mn,th Mn,th/2
experimental
[A1]
time
Mn
Mw/ [chain]
Mn (mol‚L-1
)
number of
spins of raw
(mol‚L-1
)
(h) convn (g‚mol-1) (g‚mol-1) (g‚mol-1
)
polymer solutiona
0.016
0.5 0.18
1.0 0.32
1.5 0.40
2.5 0.54
5.0 0.65
14.0 0.79
0.5 0.19
1.0 0.29
1.5 0.37
2.5 0.47
5.0 0.63
14.0 0.77
10300
17900
22800
30600
36700
45000
17300
26100
33200
42400
57000
70400
5150
8950
10200 1.58
15300 1.43
21300 1.28
23000 1.42
23400 1.45
19100 1.60
16800 1.50
23700 1.43
30400 1.35
35500 1.38
38000 1.45
31500 1.55
0.016
0.019
0.017
0.021
0.025
0.039
0.010
0.011
0.011
0.012
0.015
0.022
(10-4 mol‚L-1
)
concn of
11400
15300
18350
22500
8650
13050
16600
21200
28500
35200
bini- time
[chain]
before
oxidn
after
oxidn
hydroxylamine
troxide (h) convn (mol‚L-1
)
(mol‚L-1
)
a,b
B1
0.5 0.18
1.0 0.34
1.5 0.48
2.5 0.63
5.0 0.73
14.0 0.95
2.5 0.12
5.0 0.29
14.0 0.74
2.5 0.53
5.0 0.72
14.0 0.88
2.5 0.53
5.0 0.76
14.0 0.86
0.018
0.022
0.024
0.025
0.032
0.032
0.015
0.019
0.021
0.022
0.026
0.031
0.023
0.026
0.031
1.4
3.2
3.7
4.9
5.9
6.8
1.1
2.7
5.1
3.6
9.5
12.2
10.3
10.3
9.9
2.6
5.1
8.5
2.2
6.3
8.5
5.4
4.4
3.1
1.5
2.4
3.4
0.010
B1*
B4
change was observed in the hyperfine splitting (aN ) 15.4 G)
of the three-line spectra, indicating an isotropic motion and
thus validating the previous calculation of τc.20 The above 10-
fold increase of the smaller τc with the time of heating may be
attributed either to an increase of the (microscopic) viscosity,
the volume of the species with the smaller τc being constant,
or to a small increase of this volume, the viscosity being
constant, or to a simultaneous occurrence of both changes. In
any case, the approximately 200-fold increase of the larger τc
has to be attributed mainly to an increase in the volume of
the corresponding molecule.21
8.7
12.8
10.2
5.4
7.8
10.9
B5
a
Calculated with respect to the reaction medium volume.
Calculated from the difference between the spin number of the
b
raw polymer before and after oxidation.
Kin etic Stu d y of th e Hom olytic Dissocia tion of A1 in
th e P r esen ce of Ga lvin oxyl. A procedure similar to that in
previous literature was used.22 A large excess of galvinoxyl
radical (1 × 10-3 mol‚L-1) was added to a solution of 2 × 10-5
mol‚L-1 of A1 in tert-butylbenzene, to instantaneously scav-
enge all the carbon-centered radicals formed upon alkoxyamine
dissociation. The solution was then introduced into an ESR
tube, which was degassed by three freeze-pump-thaw cycles
and sealed off under vacuum. The tube was then heated to
130 °C, and the concentration of nitroxide radicals produced
during the course of the reaction was monitored by ESR: the
evolution of the peak at low field of the nitroxide, which
distinguished itself from the large peak of galvinoxyl, was
followed and measured.
P olym er Oxid a tion . The purified polymers (50 mg) (see
Table 3) obtained in the bulk polymerization were dissolved
in CH2Cl2 (10 mL) and a large excess of PbO2 was added. The
mixture was stirred at room temperature and the oxidation
was monitored by ESR. When the intensity of the radical
signal reached a plateau, the number of spin of the solution s
was measured and converted to the number of spin of the
polymerization medium S after oxidation. The ESR signal of
the purified polymer samples (in CH2Cl2 solution) before and
after oxidation was characteristic of that of a mononitroxide,
and after oxidation the intensity of the ESR signal increased
(see Figure 2).
P olym er Hea tin g. The purified polystyrene (10 mg), placed
in a tube open to air, was heated at 130 °C in a thermostated
oven for 24 h. The polymer was then dissolved in THF (2 mL)
for measurement of the molar mass by SEC.
Kin etic Stu d y of Hom olytic Dissocia tion of P NNP
Ch a in s in th e P r esen ce of Ga lvin oxyl. Two PNNP poly-
styrenes (P 1 and P 2) were prepared in bulk at 130 °C, using
BPO (0.030 mol‚L-1) as the initiator and B1 (0.018 mol‚L-1
;

Macromolecules, Vol. 35, No. 6, 2002
Unique Behavior of Nitroxide Biradicals 2309
F igu r e 1. ESR spectra of nitroxide radicals formed at
different time intervals in the bulk polymerization of styrene
at 130 °C, in the presence of the bisalkoxyamine A1 (0.02 mol‚
L-1): (a) 15 min; (b) 2.5 h; (c) 5 h; (d) 10 h.
F igu r e 3. ln[M]0/[M] vs time for the bulk polymerizations of
styrene at 130 °C in the presence of the bisalkoxyamine A1.
Equation of calculated curve taken from Fukuda et al.24
applied for thermal auto-initiation, with third-order kinetics
2
3
with respect to monomer. ln[M]0/[M] ) /3 ln(1 + /2A0t) with
A0 ) 1.3 × 10-4 s-1 (to fit the experimental data).
Resu lts a n d Discu ssion
1. Bu lk P olym er iza tion s of Styr en e In itia ted
eith er by BP O in th e P r esen ce of th e Va r iou s
Bin itr oxid es or by th e Bisa lk oxya m in e A1. Five
biradicals with their nitroxide groups having an envi-
ronment similar to that of TEMPO were chosen and
were thus expected to exhibit similar reactivity. The
polymerizations were performed in bulk at 130 or 110
°C using the experimental conditions described in the
literature for TEMPO-mediated polymerizations, either
with BPO as a radical initiator in the presence of
various binitroxides3 or initiated by A1 as a unimolecu-
lar initiator.4,5 As shown in Tables 1 and 2, the bini-
troxides examined behaved in a similar way, regardless
of their structure. Some typical features of nitroxide-
mediated controlled polymerizations were displayed,
such as slow polymerization, increase in Mn with
monomer conversion and polydispersity indices below
1.5 (at least in the early stage of the polymerization).
Polymerization kinetics were slower at 110 °C than at
130 °C. At the higher temperature, the same kinetics
were observed for the various binitroxides and for both
A1 concentrations (Figure 3), indicating the occurrence
of thermal auto-initiation and its strong influence on
the concentration of propagating radicals as demon-
strated for TEMPO-mediated CRP of styrene in bulk.24
Contrary to conventional mononitroxide-mediated po-
lymerizations, the purified polymer samples were para-
magnetic and their ESR spectra in THF or CH2Cl2
solvent were characteristic of that of a mononitroxide
in fluid solution (τc < 10-11 s) (Table 3 and Figure 2).
This implies first that the binitroxides were involved
in the polymerization reaction and second that part of
them remained attached to polymer chains by one side
only. The concentration of chains increased throughout
the reaction (this effect being more pronounced at 130
°C, see Table 1) in such a way that it could not be
explained by thermal auto-initiation only (thermal auto-
initiation of styrene in the absence of added nitroxide
led to the formation of no more than 0.0022 mol‚L-1
polymer chains, after 14 h in bulk at 130 °C).25 As a
consequence, the polydispersity indices also increased,
instead of progressively decreasing. The SEC traces
F igu r e 2. ESR signal of a purified polymer sample in CH2-
Cl2 solution, polystyrene prepared in bulk at 130 °C (0.5 h
polymerization time), using BPO as the initiator and B1 as
the mediator: (a) before and (b) after oxidation.
[NO•] ) 0.036 mol‚L-1) as the mediator. To avoid chain-end
decomposition, polymerization was stopped at low monomer
conversion, namely 5.1% after 20 min for P 1 and 7.6% after
40 min for P 2. The polymers were recovered, purified, and
analyzed as described above. The number average molar mass
of P 1 was Mn ) 3840 g‚mol-1, with Mw/Mn ) 1.26; for P 2, Mn
) 5230 g‚mol-1 and Mw/Mn ) 1.24. Dissociation of the
alkoxyamine end group was monitored by ESR using the same
procedure as for A1.
Ir r ever sible Decom p osition of th e Bisa lk oxya m in e A1
a n d of th e Mon oa lk oxya m in e S-TEMP O. Thermal decom-
position of bisalkoxyamine A1 and alkoxyamine S-TEMPO was
studied by 1H NMR, following the same procedure for both
compounds. A solution of the alkoxyamine in benzene (0.06
mol‚L-1) was introduced in a NMR tube. The tube was
thoroughly degassed by three freeze-pump-thaw cycles and
sealed off under vacuum. It was then immersed into an oil
bath at 130 °C. At regular time intervals, the tube was cooled
to room temperature and 1H NMR analysis was performed.
The precise protocol has already been reported.23 For the
decomposition of S-TEMPO, two byproducts, styrene and the
N-hydroxylamine derivative of TEMPO, were formed, as was
previously observed in the absence of oxygen.23 With the
heating time, the peaks corresponding to the three vinylic
protons of styrene at δ 5.3 (d), 5.7 (d), and 6.7 (dd) ppm and
the peak corresponding to the twelve methyl protons of the
hydroxylamine of TEMPO at δ 1.3 ppm (s) increased in
intensity, while the peak of the methine proton of S-TEMPO
at δ 5.0 (q) ppm decreased in intensity. For decomposition of
the bisalkoxyamine A1, the spectrum was much more com-
plicated; however, it clearly showed an increase of the vinylic
styrene protons at δ 5.3, 5.7, and 6.7 ppm, as well as a decrease
of the methine protons of the alkoxyamine at δ 3.9 (m) and
4.9 (m) ppm. Although the production of styrene and hydroxyl-
amine should occur simultaneously and in the same quantity,
the corresponding hydroxylamine could not be detected with
certainty, probably because its spectrum was too close to the
alicyclic part of the alkoxyamine.

2310 Huang et al.
Macromolecules, Vol. 35, No. 6, 2002
F igu r e 5. Mn vs monomer conversion for bulk polymerizations
of styrene at 130 °C in the presence of BPO and B1, B2, or
B3 nitroxide biradicals ([BPO] ) 0.030 mol‚L-1; [NO•] ) 2 [B1]
) 2 [B2] ) 2 [B3] ) 0.036 mol‚L-1).
F igu r e 4. SEC traces (differential refractive index) of raw
(s) and heated (oooooo) polymers obtained from bulk polymer-
ization of styrene at 130 °C in the presence of BPO and B3
nitroxide biradical ([BPO] ) 0.030 mol‚L-1; [NO•] ) 2 [B3] )
0.036 mol‚L-1): (a) conversion ) 43%; (b) conversion ) 64%,
(c) conversion ) 85%.
were particularly unusual, as illustrated in Figure 4.
Indeed, while the overall peak shifted with conversion
toward the lower elution volumes (larger molar masses)
a bimodal distribution was observed. The molar mass
values at the maximum of the two peaks were system-
atically in a ratio of 2 (see Table 1). However, the
proportions of both peaks changed dramatically with the
progress of monomer conversion: the relative area of
the low molar mass peak increased whereas that of the
high molar mass peak decreased. As a result, the Mn
vs conversion displayed in Figures 5 and 6 was not
linear. Initially, the Mn values followed the theoretical
line corresponding to polymers with two growing chains
attached to a central bisalkoxyamine core. Then the Mn
values steadily deviated from linearity to reach another
theoretical line referred to as Mn/2 and corresponding
to branches that have disconnected from the core. After
the dried polymer was heated at 130 °C for 24 h, the
SEC traces considerably changed to display a single
peak, which perfectly overlayed the lower molar mass
one in the initial distribution (Figure 4). Moreover, Mn
of such modified polymer increased linearly with mono-
mer conversion, following the Mn/2 line (Figure 5).
Indeed, when the polymer was heated to 130 °C, the
alkoxyamine C-O bonds in the polymer chains were
broken (producing unsaturation or ketone end group):
F igu r e 6. Mn vs monomer conversion for bulk polymerizations
of styrene at 130 °C in the presence of bisalkoxyamine A1.
distribution of the SEC traces, showed very clearly that
the alkoxyamine bonds were not perfectly stable during
the course of the polymerization but underwent side
reaction leading to irreversible arm separation.
Two possible reactions can account for separation of
the arms. The first one (reversible, PNNP h PNN• +
P•) is the homolytic cleavage of one alkoxyamine bond
leading to a propagating chain and a nitroxide termi-
nated polymer. The second one (irreversible, PNNP f
PNNH + Pd) is the decomposition of one alkoxyamine
bond into an unsaturated polymer (dead) and a macro-
molecular hydroxylamine (still living, because the other
alkoxyamine function is still present). The first reaction
necessarily exits, as it allows polymerization to proceed.
Moreover, the concentration of free nitroxide was rather
high compared to TEMPO-mediated CRP of styrene (see
Table 3), in the range 10-4-10-3 mol‚L-1 instead of
10-5-10-4 mol‚L-1 24
Such a result was also observed
.
for the polymeric nitroxides obtained by copolymerizing
styrene with a nitroxide-bearing comonomer.13 However,
it was not sufficiently high (< 5 mol % of the total chain
concentration) to completely explain the large propor-
26
the two-arm macromolecules broke in the middle,
supporting both segments having the same length owing
to fast exchange. This result, together with the bimodal

Macromolecules, Vol. 35, No. 6, 2002
Unique Behavior of Nitroxide Biradicals 2311
Sch em e 3. Rea ction Sch em e for Bu lk P olym er iza tion
of Styr en e In itia ted by th e Bisa lk oxya m in e A1
tion of divided chains. Thus the second mechanism was
also supposed to occur to a large extent at 130 °C. At a
lower temperature, namely 110 °C, a better control of
the two-arm structure was observed, since experimental
molar masses remained close to the theoretical values,
even at 74% conversion (Table 1). To ascertain the
occurrence of this second reaction, the polymer was
analyzed by ESR after mild oxidation with PbO2. If
present, the hydroxylamines would be oxidized into
nitroxides, and the number of spin would increase after
oxidation, which was indeed observed (Figure 2; Table
3).
On the basis of all the experimental observations, the
major reactions involved are summarized in Scheme 3.
They lead to a mixture of polymer chains mainly
composed of PNNP, PNNH (turned into PNN•, after
mild oxidation), PNN•, and Pd in proportions that vary
with conversion. To fully understand the mechanisms
that govern the behavior of the studied nitroxide bi-
radicals in the CRP of styrene, a thorough kinetic study
of the main reactions depicted in Scheme 3 (namely the
activation-deactivation equilibrium of the polymeric
species and the alkoxyamine decomposition) has been
undertaken and is presented.
2. Deter m in a tion of th e Ra te Con sta n t of Dis-
socia tion , kd ,0, of th e Bisa lk oxya m in e A1. The
kinetics of the homolytic cleavage of A1 were monitored
by real-time quantitative ESR analysis of the concen-
tration of released nitroxide. The dissociation reaction
was performed at 130 °C in the presence of galvinoxyl
and the concentration of nitroxide vs time is plotted in
Figure 7. Providing that the carbon-centered radicals
produced were instantaneously trapped by the galvi-
noxyl radical in order to prevent the back reaction, the
initial rate of appearance of the nitroxide gave the value
of kd,0 of reaction 1 in Scheme 3, according to eq III.
d[SNN•]
(
)
dt
[A1]0
t)0
kd,0
)
(III)
Although the reaction was allowed to proceed for a
long period of time, for a better accuracy of the result
only the initial slope was used to determine kd,0. The
structure of the nitroxide formed evolved progressively
•
from SNN• (reaction 1) to NN• (reaction 2). The latter
exhibited a very broad ESR spectrum which was not
taken into consideration for the quantitative analysis
(only the peak at low field of the SNN• nitroxide was
followed and measured). In addition, B1, B2 and B3
•
nitroxide biradicals NN• were quite unstable at 130 °C:
27
B1 completely decomposed within 2 h, and B2 and
B3, within 4 h. Hence, only the concentration of the
monoradical SNN• was quantitatively monitored in the
applied procedure. Thus, the measured nitroxide con-
centration did not reach the maximum value of 4 × 10-5
mol‚L-1 (corresponding to twice the initial concentration
of A1), but passed through a maximum at a lower value
and then continuously decreased. The calculated value
of kd,0 at 130 °C was 6.6 × 10-4 s-1. It can be compared
with data reported for the analogous monoalkoxy-
amine S-TEMPO. Dissociation of the latter was
studied by Skene et al.28 in cyclohexanol in a temper-
ature range of 75-120 °C: the activation energy was
Ea ) 128 kJ ‚mol-1, the preexponential factor was A )
kd,130°C ) 1.4 × 10-3 s-1 in tert-butylbenzene. Another
alkoxyamine with structure very close to S-TEMPO, the
2-tert-butoxy-1-phenyl-1-(1-oxy-2,2,6,6-tetramethylpip-
eridinyl)ethane, was examined by Bon et al.30 in toluene
5.0 × 1013 s-1, and calculated kd,130°C ) 1.1 × 10-3 s-1
.
It was also studied by Marque et al.29 who obtained

2312 Huang et al.
Macromolecules, Vol. 35, No. 6, 2002
F igu r e 8. ln([A1]0/[PNNP]t) vs time for the bulk polymeriza-
tions of styrene at 130 °C in the presence of bisalkoxyamine
A1. The slope gives the apparent first-order rate constant kdec
F igu r e 7. Decomposition of bisalkoxyamine A1 at 130 °C in
tert-butylbenzene solutions, in the presence of galvinoxyl. The
formation of SNN• nitroxide radicals is followed by ESR
analysis. [A1] ) 2 × 10-5 mol‚L-1; [galvinoxyl] ) 1 × 10-3
) 3.8 × 10-5 s-1
.
toluene solution), which was explained by the deactivat-
ing effect of the carbonyl group in the molecule. Con-
cerning P-TEMPO, the same authors found kdec ) 8.4
× 10-6 s-1 at 130 °C in toluene solution (t1/2 ) 22.9 h).
Again the rate constant of decomposition was found to
be below that of S-TEMPO, which was not clearly
explained. The decomposition of P-TEMPO in styrene
during polymerization was also studied by Zhu et al.35
and Hawker et al.36 They found that alkoxyamine end
functionality decreased with increasing monomer con-
version and target molar mass. An end group purity
larger than 90% was obtained for low molar mass
polystyrene only (Mn < 10 000 g‚mol-1). In the poly-
merization of styrene initiated by a styryl-based alkoxy-
amine with a chromophore attached to the nitroxide
group (TEMPO type), the chain-end purity was ap-
proximately 60% after 16 h polymerization time at 120
°C.35 This leads to a rate constant of decomposition of
approximately 9 × 10-6 s-1, which is not very different
from the value obtained in an inert solvent.
In this work, the decomposition rate of polystyryl-
based bisalkoxyamine PNNP was not studied directly
from NMR analysis in an inert solvent but was deduced
from the proportion of dead chains formed during the
course of the polymerization. As mentioned in the first
paragraph, hydroxylamine simultaneously formed, the
concentration of which was calculated from ESR mea-
surements after mild oxidation of the polymer samples
(see Table 3). The concentration went however through
a maximum, indicating the slow degradation of the
hydroxylamine under the polymerization conditions.37
Therefore, this technique was not appropriate for an
accurate determination of the rate constant of formation
of the hydroxylamine, and hence kdec. Assuming that
the chain-end decomposition was the principal chain-
breaking reaction in the system (as also supposed and
well argued by Ohno et al.),34 the extent of this reaction
could be estimated from the concentration of chains
derived from conversion and Mn data (see Table 2,
polymerizations initiated with A1). The first step of the
decomposition reaction producing PNNH and Pd led to
an increase in the chain concentration, with [Pd]t )
[chains]t - [A1]t)0 and [PNNP]t ) [A1]0 - [Pd]t )
2[A1]t)0 - [chains]t. Considering decomposition as an
apparent first-order reaction, ln([A1]t)0/[PNNP]t) is
plotted vs time in Figure 8 for two A1 initial concentra-
tions, namely 0.010 and 0.016 mol‚L-1. The slope gave
kdec ) 3.8 × 10-5 s-1 at 130 °C (t1/2 ) 5.1 h; 85 mol % of
chains have decomposed within 14 h), under polymer-
ization conditions, i.e., in styrene solution. This value
mol‚L-1
.
in a temperature range of 60-100 °C. They reported
Ea ) 133 kJ ‚mol-1 and A ) 1.1 × 1014 s-1 and calculated
kd, 130 °C ) 6.0 × 10-4 s-1. In the case of the bis-
alkoxyamine A1, the measured kd,0 value at 130 °C is
not very different from those obtained for similar
monoalkoxyamines, indicating no effect of the binitrox-
ide structure on the dissociation reaction.
3. Deter m in a tion of th e Ra te Con sta n t of Dis-
socia t ion , k d ,0, of P olym er ic Bisa lk oxya m in es
P NNP . Concerning the TEMPO-based polystyryl alkoxy-
amine (P-TEMPO), Goto et al.31 determined Ea ) 124
kJ ‚mol-1, A ) 3.0 × 1013 s-1 and kd,130°C ) 2.5 × 10-3
s-1 while Bon et al.30 reported Ea ) 141 kJ ‚mol-1, A )
1.0 × 1016 s-1 and kd,130°C ) 5.6 × 10-3 s-1. These
observed 2- to 9-fold larger kd values for the polymeric
species with respect to the small molecules were as-
cribed to a larger steric strain on the breaking bond,29,30
but the difference was generally considered as rather
marginal.29,31,32 To compare with the styryl-based bis-
alkoxyamine A1, the dissociation of two PNNP poly-
meric bisalkoxyamines P 1 and P 2 was examined at 130
°C, using the same experimental procedure. The rate
constant of dissociation, kd,0, was 1.5 × 10-3 s-1 for P 1
and 1.4 × 10-3 s-1 for P 2. The values were only slightly
larger than that determined for A1, indicating no strong
effect of the chain length. In addition, the rate constant
of dissociation was not very different from that of
P-TEMPO, indicating that the presence of a polymeric
chain attached to the leaving nitroxide had no signifi-
cant influence on the homolytic cleavage of the alkoxy-
amine bond.
4. Det er m in a t ion of t h e Ap p a r en t F ir st -Or d er
Ra te Con sta n t of Decom p osition (kd ec). The proton
NMR analysis of the bisalkoxyamine A1 in benzene
solution at 130 °C showed the existence of a decomposi-
tion reaction leading to styrene and supposedly hy-
droxylamine in the absence of oxygen in the NMR tube.
When oxygen was present, small fractions of acetophen-
one and 1-phenyl-1-ethanol were also detected. The
results were not different from those also observed for
the S-TEMPO alkoxyamine and already described in the
literature.23 A kinetic study concerning the decomposi-
tion of S-TEMPO alkoxyamine in toluene at 130 °C,
gave kdec ) 5.0 × 10-5 s-1 (t1/2 ) 3.9 h).33 Surprisingly,
the rate constant found by Ohno34 for the 2-benzoyloxy-
1-phenylethyl TEMPO-based alkoxyamine was 1 order
of magnitude lower (kdec ) 2.1 × 10-6 s-1 at 130 °C in

Macromolecules, Vol. 35, No. 6, 2002
Unique Behavior of Nitroxide Biradicals 2313
is four to five times larger than the kdec previously
reported for P-TEMPO in toluene at 130 °C,34 or than
the kdec calculated from published polymerization data.35
The strong discrepancy with TEMPO-mediated poly-
merization of styrene might be related to the unique
structure of the nitroxide counterradical, PNN•. Indeed,
as long as the chains grow, overall size of this nitroxide
increases accordingly. This effect will be discussed more
thoroughly in the last part of this article, in light of the
full set of rate constants.
5. Deter m in a tion of th e Equ ilibr iu m Con sta n ts
K a n d K′ of Rea ction s 5 a n d 6 a t 130 °C. The
binitroxide-mediated polymerization of styrene cannot
be regarded as a simple system, regulated by only one
activation-deactivation equilibrium. Indeed, when mono-
mer conversion progresses and PNNP chains decompose
into HNNP (and eventually X-NP, as the hydroxylamine
group is not stable), the activation-deactivation equili-
brated reactions 5 and 6 occur simultaneously. Then,
two different nitroxides, PNN• and HNN• in continu-
ously changing proportions, actually control the poly-
merization. As shown in Scheme 3, the probability of
F igu r e 9. Overall concentration of nitroxide, [N•] ) [PNN•]
+ [HNN•], vs time followed by in situ ESR analysis for bulk
polymerizations of styrene at 130 °C in the presence of
bisalkoxyamine A1 at two initial concentrations.
•
formation of the binitroxide NN• is too low to consider
this species in the kinetics. The way both equilibrium
relationships can be written depends on the very mech-
anism of the decomposition reaction (see appendix).
However, it is still not clear in the literature whether
such irreversible decomposition follows a true first order
kinetics (reactions 10 and 12; Scheme 3) or a second
order one (i.e., hydrogen transfer from propagating
radical to nitroxide, reactions 9 and 11) because it was
demonstrated that they cannot be kinetically differenti-
ated.38 In the case of a first-order decomposition reac-
tion, kdec is a true first-order rate constant, while in the
case of a second-order decomposition reaction, the rate
constant is ktr, and an apparent first-order rate constant
kdec can be calculated as kdec ) ktrkd/(kc + ktr). When ktr
is small with respect to kc (in such a case kdec becomes
Kktr) (see Appendix), then both mechanisms lead to the
same formula for the overall nitroxide concentration (eq
IV), from which an apparent equilibrium constant can
F igu r e 10. [P•] vs time for the bulk polymerizations of styrene
at 130 °C in the presence of the bisalkoxyamine A1. Calculated
curve (derivative of the curve given in Figure 3): [P•] ) A0/
kp(1 + 3/2A0t) with kp ) 2600 L‚mol-1‚s-1 39 and A0 ) 1.3 ×
10-4 s-1
.
time plot given in Figure 3, using the propagation rate
constant kp ) 2600 L‚mol-1‚s-1 at 130 °C.39 To facilitate
the calculation and take advantage of the large number
of data points of [N•], the [P•] vs time plot was fitted
with a mathematical curve (the same for both initial
concentrations of initiator, which had no significant
effect on the rate) from which [P•] was calculated for
each time an experimental value of [N•] was available.
The apparent equilibrium constant Kapp ) [N•][P•]/[A1]0
was plotted as a function of polymerization time for both
A1 initial concentrations 0.010 and 0.020 M (Figure
11a). Interestingly, Kapp continuously decreased with
polymerization time, from 3.0 × 10-10 mol‚L-1 (corre-
sponding to K in eq A9 or to kd/(kc + ktr) in eq A24) to a
value around 1 × 10-10 mol‚L-1 (corresponding to K′ or
to k′d/(k′c + k′tr) at infinite time). Considering eq A9,
one can calculate K ) 3.0 × 10-10 mol‚L-1 and kc ) kd/K
) 4.7 × 106 L‚mol-1‚s-1 using kd ) 1.4 × 10-3 s-1 for
PNNP. With equation (A24), the ratio kd/(kc + ktr) equals
3.0 × 10-10 mol‚L-1; with kd ) 1.4 × 10-3 s-1, one has
kc + ktr ) 4.7 × 106 L‚mol-1‚s-1; and finally, with eq
A15 and kdec ) 3.8 × 10-5 s-1, it is possible to calculate
ktr ) 1.3 × 105 L‚mol-1‚s-1; kc ) 4.6 × 106 L‚mol-1‚s-1
and K ) 3.0 × 10-10 mol‚L-1. In the latter case, equation
(A15) simplifies into kdec ) Kktr and both situations (first
or second-order decomposition reaction) lead to the same
results for K and kc, respectively. All the data are
gathered in Table 4 with those already reported for
P-TEMPO. It is quite interesting to note that K is very
[N•] ) [PNN•] + [HNN•] )
K[PNNP]t + K′[HNNP]t
)
[P•]
(K - K′)[A1]0 exp(-kdect) + K′[A1]0
(IV)
[P•]
be deduced (eq V).
[N•][P•]
Kapp
)
) (K - K′) exp(-kdect) + K′ (V)
[A1]0
Variation of the concentration of free nitroxide with
polymerization time was determined in situ by real-time
ESR analysis (see Experimental Section) for two A1
initial concentrations, namely, 0.010 and 0.020 mol‚L-1
at 130 °C (Figure 9). Measurements gave an overall
concentration, [N•] ) [PNN•] + [HNN•]. This was
supported by the ESR spectra showing two mononitrox-
ides with different rotational correlation time; in addi-
tion two different mononitroxides were detected after
purification of the polymer, respectively in the recovered
polymer (PNN•) and in the filtrate (HNN• or a deriva-
tive). The experimental concentration of P• (Figure 10)
was calculated from the tangent of the ln[M]0/[M] vs

2314 Huang et al.
Macromolecules, Vol. 35, No. 6, 2002
two macromolecules of the same size, namely PNN• and
P•, and, also admitted for kt,40 kc might be chain-length
dependent (kc would decrease and K would increase with
the increase of chain length, and hence with conversion).
6. Effect on P olym er iza tion Kin etics of th e
Existen ce of Tw o Sim u lta n eou s Equ ilibr ia . It was
shown in Figure 3 that the polymerization rate was
almost independent of the initial concentration of A1.
This was not surprising as polymerization was con-
ducted at 130 °C, where thermal auto-initiation of
styrene cannot be neglected. However, the concentration
of P• significantly decreased throughout the polymeri-
zation reaction, whereas Fukuda demonstrated that
only a slight decrease could be observed in TEMPO-
mediated polymerization of styrene, owing to the de-
crease of Ri with monomer consumption (reaction 7,
Scheme 3).24,41 The pronounced decrease in our system
is concomitant with an increase in the concentration of
free nitroxide (Figure 9). Such a continuous increase in
the concentration of the persistent radical with time is
the signature of the so-called persistent radical effect
(PRE),42 which is usually concealed when only thermal
auto-initiation governs the kinetics.24,32 Indeed, the
continuous thermal generation of radicals contributes
to the creation of new chains, which in turn are capped
by free nitroxide molecules released upon termination
of propagating radical. This effect levels off the nitroxide
concentration, while the natural trend would be a
continuous built up. In our system, however, the be-
havior is somewhat different. Figure 9 illustrates the
continuous increase in nitroxide concentration with
polymerization time. Furthermore, it proves it was
dependent on the initial concentration of A1. This could
be anticipated from eq IV which points out that a direct
proportionality should exist between [N•] and the con-
centration of living chains, providing that [P•] is inde-
pendent of this concentration. Actually, the concentra-
tion of released nitroxide approximately doubled when
the initial concentration of A1 was multiplied by 2. All
these results show quite clearly that the kinetics of our
system can be perfectly described neither by the PRE
(the polymerization rate does not depend on [A1]0, [N•]
F igu r e 11. Apparent equilibrium constant Kapp as a function
of time (a) and of exp(-kdect) (b) for the bulk polymerization
of styrene at 130 °C initiated by bisalkoxyamine A1 at two
initial concentrations (0.010 and 0.020 M), using kdec ) 3.8 ×
10-5 s-1
.
Ta ble 4. Kin etic P a r a m eter s a t 130 °C for P NNP a n d
P -TEMP O Alk oxya m in es
P-TEMPO
PNNP
kd (s-1
)
2.5 × 10-3, ref 31
5.6 × 10-3, ref 30
1.2 × 108, ref 31
2.1 × 10-11, ref 31
8.4 × 10-6, ref 34
kdec/K ) 4.0 × 105
1.4 × 10-3
kc (L‚mol-1‚s-1
)
4.6 × 106
3.0 × 10-10
3.8 × 10-5
1.3 × 105
K (mol‚L-1
kdec (s-1
ktr (L‚mol-1‚s-1
)
)
is not proportional to [A1]0 )
2/3 42 nor by the simple steady
)
state induced by thermal auto-initiation ([P•] and [N•]
are not constant). It is also noticeable in Figure 9 and
Table 3 that the concentration of released nitroxide was
approximately 1 order of magnitude larger than that
usually observed in TEMPO-mediated polymerization
of styrene.24 This last result can be explained by a larger
value of the equilibrium constant K of reaction 5, with
respect to that observed in the P-TEMPO system.
large in this system and kc is very small in comparison
with the values reported for P-TEMPO (K ) 2.1 × 10-11
mol‚L-1 and kc ) 1.2 × 108 L‚mol-1‚s-1).31 This result
can be explained by steric hindrance of the nitroxide
counterradical PNN•, which is a polymeric species. The
same trend was previously observed for the polymeric
nitroxides, which were shown to be less efficient at
trapping propagating radicals than the classical mono-
nitroxides.13
When plotting Kapp as a function of exp(-kdect), one
cannot get a perfectly linear evolution (Figure 11b) as
anticipated by eq V. This makes it difficult to accurately
determine K′ according to this equation. The trend
however indicates that K′ is quite smaller than K, which
is actually not unexpected. Indeed, equilibrium 6 con-
cerns the small nitroxide HNN•, which should not
behave very differently from TEMPO, as it has a similar
structure. The imperfect linearity of Kapp with exp(-
kdect) might be the consequence of a change in the value
of K (coming from a change in kc) with the progress of
monomer conversion and the increase in viscosity.
Indeed, recombination is a bimolecular reaction between
However, as the apparent equilibrium constant, Kapp
,
decreases from K to K′ throughout the reaction owing
to the existence of equilibrium 6 (Scheme 3), the system
is shifted toward the formation of a larger proportion
of dormant species. The buildup of free nitroxide is not
as large as it would have been in a simple system with
a single activation-deactivation equilibrium. The be-
havior of our system is between that observed for the
polymerization of styrene mediated by the acyclic ni-
troxide SG1 (N-tert-butyl-N-(1-diethylphosphono-2,2-
dimethylpropyl) nitroxide), for which the kinetics was
shown to be regulated by the PRE owing to a large
equilibrium constant43 and TEMPO-mediated polymer-
ization, the kinetics of which is governed by thermal
auto-initiation.24

Macromolecules, Vol. 35, No. 6, 2002
Unique Behavior of Nitroxide Biradicals 2315
7. Discu ssion on th e Mech a n ism of th e Decom -
p osition Rea ction of th e Alk oxya m in e En d Gr ou p .
In the binitroxide-mediated polymerization of styrene
studied in this work, one of the main features is the
fast decomposition of the alkoxyamine end group, as
compared with TEMPO-mediated polymerization. From
a kinetic viewpoint, this decomposition reaction can be
seen either as a first-order reaction or as a second-order
one, without any possible distinction.38 If formation of
the dead chains followed a unimolecular mechanism,
because chemical structure of the polymeric bisalkoxy-
amines studied is similar to that of P-TEMPO, the rate
constants of decomposition, kdec, would not be so differ-
ent. For comparison, the rate constants of homolytic
dissociation, kd, for both alkoxyamines were quite
similar. If, in contrast, decomposition is a bimolecular
reaction between propagating radical and nitroxide,
then the reason for different kinetics can be more easily
explained. Indeed, the large activation-deactivation
equilibrium constant of reaction 5 allows in the system
the simultaneous existence of a high concentration of
both P• and PNN•. Thus, even if the rate constant ktr of
the hydrogen abstraction reaction is not different in this
system from that observed in the P-TEMPO system (it
is even lower; see Table 4), the apparent first-order rate
constant kdec, which equals ktr.K can be significantly
larger if K is larger. At 110 °C, the two-arm structure
is more stable, because of a lower value of ktrK.
enhancement of this reaction in binitroxide-mediated
CRP of styrene was assigned to the large concentrations
of both propagating and persistent radicals. For these
reasons, the use of multinitroxides in radical polymer-
ization does not allow the synthesis of well-defined
multiarm macromolecules at large monomer conversion.
The quality of the structure can however be improved
when polymerization is performed at a temperature of
110 °C, instead of 130 °C.
Ap p en d ix. Over a ll Nitr oxid e Con cen tr a tion a n d
Ap p a r en t Activa tion -Dea ctiva tion Equ ilibr iu m
Con sta n t for Tw o Differ en t Ca ses
(i) Ca se of a F ir st-Or d er Ir r ever sible Decom p o-
sition Rea ction . Consider reactions 5, 6, 7, 8, 10, and
12 depicted in Scheme 3. The rate of evolution of the
concentration of propagating radicals is
d[P•]
) Ri + kd[PNNP] + k′d[HNNP] - (kc[PNN•] +
dt
k′c[HNN•])[P•] - kt[P•]2 (A1)
) 0
according to the kinetic treatment proposed by Fukuda
when thermal initiation takes place.24
The true first-order decomposition of the alkoxyamine
leads to the following rate and concentration equations:
Con clu sion
The use of binitroxides, with both radical sites of
similar reactivity, as mediator in the controlled radical
polymerization of styrene was examined and the rate
constants of the variously involved reactions were
determined at 130 °C. The typical features of a con-
trolled polymerization were observed in the early stage
of the polymerization, leading to the formation of two-
arm macromolecules containing the binitroxide at the
core. Then a deviation occurred owing to continuous
irreversible termination leading to unsaturated polymer
and hydroxylamine. This behavior was particularly
observed in the SEC traces, with a bimodal distribution
assigned to a mixture of two-arm and one-arm macro-
molecules. After irreversible separation, the one-arm
macromolecules formed were of two kinds: Pd (dead
chains with terminal unsaturation) and PNNH (living
chains with a hydroxylamine attached to the nitroxide
moiety). The apparent first-order rate constant of the
decomposition reaction was much larger than in TEMPO-
mediated polymerization of styrene. In contrast, the rate
constants of the reversible dissociation of the low molar
mass model bisalkoxyamine (SNNS, A1) and of the
polymeric bisalkoxyamines (PNNP) were not very dif-
ferent from each other, and not very different from those
determined for S-TEMPO or P-TEMPO, indicating no
strong effect of chain length, neither from the alkyl
radical (P•), nor from the nitroxide (PNN•). In contrast,
the rate constant of recombination between a propagat-
ing radical P• and a macromolecular counterradical
PNN• was shown to be much lower than with the
modified counterradical HNN• and lower than with
TEMPO, indicating a strong effect of the nitroxide size.
This resulted in a large activation-deactivation equi-
librium constant and a high concentration of free
nitroxide in the system. As a consequence, chain-end
decomposition was supposed to be a second-order reac-
tion, i.e., â-hydrogen transfer from P• to PNN•, rather
than a first-order decomposition of PNNP. The strong
d[PNNP]
) -kdec[PNNP]t
(A2)
dt
[PNNP]t ) [A1]0 exp(-kdect)
(with [A1]0 ) [PNNP]0) (A3)
[HNNP]t ) [A1]0 - [PNNP]t )
[A1]0 (1 - exp(- kdect)) (A4)
For simplicity, the decomposition reaction of HNNP into
HNNH (reaction 12) was neglected, assuming that it
would decompose very slowly, such as P-TEMPO.
Owing to thermal auto-initiation, the concentration
of propagating radicals is
Ri
[P•] )
(A5)
k
x
t
independent of the initial concentration of alkoxyamine
initiator in the system, providing that both equilibria 5
and 6 are established:
K[PNNP]t K[A1]0 exp(-kdect)
[PNN•] )
)
[P•]
[P•]
equilibrium 5 (A6)
and
K′[HNNP]t K′[A1]0(1 - exp(- kdect))
[HNN•] )
)
[P•]
[P•]
equilibrium 6 (A7)
The overall concentration of free nitroxide reads then

2316 Huang et al.
Macromolecules, Vol. 35, No. 6, 2002
[N•] ) [PNN•] + [HNN•] )
As previously, for simplicity, the second term was finally
neglected, considering that the decomposition of HNNP
was sufficiently slow in the system so as not to interfere.
This gives then
K[PNNP]t + K′[HNNP]t
)
[P•]
(K - K′)[A1]0 exp(-kdect) + K′[A1]0
[HNNP]t ) [A1]0 (1 - exp(- kdect))
(A20)
(A8)
[P•]
and
An apparent equilibrium constant can be deduced:
k′d[A1]0(1 - exp(-kdect))
[HNN•]t )
Kapp ) [N•][P•]/[A1]0 ) (K - K′) exp(-kdect) + K′
(A21)
(k′c + k′tr)[P•]
(A9)
This situation does not change the concentration of
propagating radicals, which remains as given in eq A5.
The overall concentration of free nitroxide reads then
(ii) Ca se of a Secon d -Or d er Ir r ever sible Decom -
p osition Rea ction . Consider now reactions 5, 6, 7, 8,
9, and 11. The hydrogen transfer reaction affects the
rate of evolution of [P•] and the concentration of nitrox-
ide in the following way:
kd[A1]0 exp(-kdect)
[N•] ) [PNN•] + [HNN•] )
+
(kc + ktr)[P•]
d[P•]
dt
k′d[A1]0(1 - exp(-kdect))
(k′c + k′tr)[P•]
) Ri + kd[PNNP] + k′d[HNNP] - (kc + ktr)
[PNN•][P•] - (k′c + k′tr)[HNN•][P•] - kt[P•]2 ) 0
(A22)
and the apparent equilibrium constant is
(A10)
d[PNN•]
kd exp(-kdect)
) kd[PNNP] - (kc + ktr) [PNN•][P•] ) 0
Kapp ) [N•][P•]/[A1]0 )
+
dt
(kc + ktr)
k′d(1 - exp(-kdect))
(k′c + k′tr)
(A11)
kd[PNNP]t
(kc + ktr)[P•]
(A23)
[PNN•]t )
(A12)
K
K′
k′tr
K′
k′tr
Kapp
)
-
exp(-kdect) +
ktr
kc
In this case, the decomposition rate of the PNNP
alkoxyamine is a true second-order reaction, which can
be written as an apparent first-order reaction
1 +
1 +
1 +
(
)
k′c
k′c
(A24)
-ktrkd
d[PNNP]
) -ktr[PNN•][P•] )
[PNNP]t
which is the same as eq A9 when ktr is small with
respect to kc (in such a case eq A15 becomes: kdec ) Kktr)
and k′tr is small with respect to k′c.
dt
(kc + ktr)
(13)
[PNNP]t ) [A1]0 exp(- kdect)
(A14)
Refer en ces a n d Notes
with the apparent rate constant
(1) (a) Controlled Radical Polymerization; Matyjaszewski, K.,
Ed.; ACS Symposium Series 685; American Chemical Soci-
ety: Washington, DC, 1998. (b) Controlled/ Living Radical
Polymerization: Progress in ATRP, NMP, and RAFT; Maty-
jaszewski, K., Ed.; ACS Symposium Series 768, American
Chemical Society: Washington, DC, 2000.
(2) Solomon, D. H.; Rizzardo, E.; Cacioli, P. U.S. Patent 4,581,-
429, March 27, 1985.
(3) Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer,
G. K. Macromolecules 1993, 26, 2987.
ktrkd
kdec
)
(A15)
(kc + ktr)
Then, eq A12 reads as follows:
kd[A1]0 exp(-kdect)
[PNN•]t )
(A16)
(4) Hawker, C. J . J . Am. Chem. Soc. 1994, 116, 11185.
(5) Malmstro¨m, E. E.; Hawker, C. J . Macromol. Chem. Phys.
1998, 199, 923 and references therein.
(kc + ktr)[P•]
(6) (a) Benoit, D.; Grimaldi, S.; Finet, J .-P.; Tordo, P.; Fontanille,
M.; Gnanou, Y. In Controlled Radical Polymerization; Maty-
jaszewski, K., Ed.; ACS Symposium Series 685; American
Chemical Society: Washington, DC, 1998; Chapter 14, p 225.
(b) Lacroix-Desmazes, P.; Lutz, J . F.; Boutevin, B. Macromol.
Chem. Phys. 2000, 201, 662. (c) Robin, S.; Gnanou, Y. In
Controlled/ Living Radical Polymerization: Progress in ATRP,
NMP, and RAFT; Matyjaszewski, K., Ed.; ACS Symposium
Series 768; American Chemical Society: Washington, DC,
2000; Chapter 23, p 334. (d) Grimaldi, S.; Finet, J .-P.; Le
Moigne, F.; Zeghdaoui, A.; Tordo. P.; Benoit, D.; Fontanille,
M.; Gnanou, Y. Macromolecules 2000, 33, 1141. (e) Benoit,
D.; Grimaldi, S.; Robin, S.; Finet, J .-P.; Tordo, P.; Gnanou,
Y. J . Am. Chem. Soc. 2000, 122, 5929.
The evolution with time of the other nitroxide, HNN•,
is given by
d[HNN•]
) k′d[HNNP] - (k′c + k′tr)[HNN•][P•] ) 0
dt
(A17)
k′d[HNNP]t
[HNN•]t )
(A18)
(k′c + k′tr)[P•]
) ktr[PNN•][P•] - k′tr[HNN•] [P•] )
d[HNNP]
(7) (a) Benoit, D.; Chaplinski, V.; Braslau, R.; Hawker, C. J . J .
Am. Chem. Soc. 1999, 121, 3904. b) Benoit, D.; Harth, E.;
Fox, P.; Waymouth, R. M.; Hawker, C. J . Macromolecules
2000, 33, 363.
dt
ktrkd
(kc + ktr)
k′trk′d
[PNNP]t -
[HNNP]t (A19)
(k′c + k′tr)
(8) Rassat, A. Pure Appl. Chem. 1990, 62, 223.

Macromolecules, Vol. 35, No. 6, 2002
Unique Behavior of Nitroxide Biradicals 2317
(9) a) Rosanzen, E. V.; Golubev, V. A.; Neiman, M. B. Izv. Akad.
Nauk SSSR Ser. Khim. 1965, 393; Bull. Acad. Sci. USSR
Div. Chem. Sci. (Engl. Transl.) 1965, 382. (b) Dupeyre, R.
M.; Lemaire, H.; Rassat, A. J . Am. Chem. Soc. 1965, 87, 3771.
(c) Parmon, V. N.; Kokorin, A. I.; Zhidomirov, C. M. Stable
biradicals (in Russian); Nauka: Moscow, 1980.
(10) (a) Falle, H. R.; Luckurst, G. R.; Lemaire, H. R.; Mare´chal,
Y.; Rassat, A.; Rey, P. Mol. Phys. 1966, 11, 49. (b) Luckurst,
G. R. In Spin Labeling Theory and Applications; Berliner, L.
J ., Ed.; Academic Press: New York, 1976.
zation of styrene. Ph.D. Thesis dissertation, University Pierre
and Marie Curie (Paris 6), Paris, 2001.
(26) Dourges, M. A.; Charleux, B.; Vairon, J .-P.; Blais, J .-C.;
Bolbach, G.; Tabet, J .-C. Macromolecules 1999, 32, 2495.
(27) To study the stability of B1, B2, and B3, solutions 0.005 M
of the bisnitroxides in dichlorobenzene were placed in capil-
lary tubes, which were sealed and heated to the desired
temperature. Decay of the nitroxide radicals was monitored
by ESR.
(28) Skene, W. G.; Belt, S. T.; Connolly, T. J .; Hahn P.; Scaiano,
J . C. Macromolecules 1998, 31, 9103.
(29) Marque, S.; Le Mercier, C.; Tordo, P.; Fischer, H. Macromol-
ecules 2000, 33, 4403.
(30) Bon, S. A. F.; Chambard, G.; German A. L. Macromolecules
1999, 32, 8269.
(31) Goto, A.; Terauchi, T.; Fukuda, T.; Miyamoto, T. Macromol.
Rapid Commun. 1997, 18, 673.
(32) Fukuda, T.; Goto, A.; Ohno, K. Macromol. Rapid Commun.
2000, 21, 151.
(33) Li, I.; Howell, B. A.; Matyjaszewski, K.; Shigemoto, T.; Smith,
P. B.; Priddy, D. B. Macromolecules 1995, 28, 6692.
(34) Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 1997, 30,
2503.
(11) Chiarelli, R.; Novak, M. A.; Rassat, A.; Tholence, J . L. Nature
1993, 363, 147.
(12) Marx, L.; Rassat A. Angew. Chem., Int. Ed. 2000, 39, 4494.
(13) (a) Li, C.; He, J .; Li, L.; Cao, J .; Yang, Y. Macromolecules
1999, 32, 7012. (b) Tao, Y.; He, J .; Wang, Z.; Pan, J .; J iang,
H.; Chen, S.; Yang, Y. Macromolecules 2001, 34, 4742.
(14) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis;
Wiley-Interscience: New York, 1967; Vol. 1, p 135.
(15) Furniss, B. S.; Hannaford, A. J .; Smith, P. W. G.; Tatchell,
A. R. Vogel’s Textbook of Practical Organic Chemistry, 5th
ed.; Wiley & Sons: New York, 1989; p 398.
(16) Murayama, K.; Morimura, S.; Amakasu, O.; Toda, T.; Yamao,
E. Nippon Kagaku Zasshi 1969, 90, 296.
(17) (a) Rozantsev, E. G.; Golubev, V. A.; Neiman, M. B.; Kokha-
nov, J . V. Izvest. Akad. Nauk SSSR. 1965, 572; Bull. Acad.
Sci. USSR Div. Chem. Sci. (Engl. Transl.) 1965, 559. (b)
Lemaire, H.; Rassat, A.; Rey, P. Bull. Soc. Chim. Fr. 1968,
3, 886.
(35) (a) Zhu, Y.; Howell, B. A.; Priddy, D. B. Am. Chem. Soc.,
Polym. Prepr. 1997, 38 (1), 97. (b) Zhu, Y.; Li, I. Q.; Howell,
B. A.; Priddy, D. B. Am. Chem. Soc., Symp. Series 1998, 685,
214.
(36) Hawker, C. J .; Hedrick, J . L. Macromolecules 1995, 28, 2993.
(37) Thermal stability of three hydroxylamines derived respec-
tively from TEMPO, TEMPONE (4-oxo-2,2,6,6-tetrameth-
ylpiperidine-N-hydroxyl) and TEMPOL (4-hydroxy-2,2,6,6-
tetramethylpiperidine-N-hydroxyl) was studied at 130 °C
using 13C NMR in DMSO-d6 solutions. Instability was
exhibited by the continuous decrease of the hydroxylamine
peaks with respect to those of the solvent.
(18) Dao, J .; Benoit, D.; Hawker, C. J . J . Polym. Sci., Part A:
Polym. Chem. 1998, 36, 2161.
(19) (a) Howell, B. A.; Priddy, D. B.; Li, L. Q.; Smith, P. B.; Kastl,
P. E. Polym. Bull. 1996, 37, 451. (b) Hawker, C. J .; Barclay,
G. G.; Orellana, A.; Dao, J .; Devonport, W. Macromolecules
1996, 29, 5245.
(20) Martinie, J .; Michon, J .; Rassat, A. J . Am. Chem. Soc. 1975,
97, 1818 and references therein.
(38) Souaille, M.; Fischer, H. Macromolecules 2001, 34, 2830.
(39) Buback, M.; Gilbert, R. G.; Hutchinson, R. A.; Klumperman,
B.; Kuchta, F. D.; Manders, B. G.; O’Driscoll, K. F.; Russell,
G. T.; Schweer, J . Macromol. Chem. Phys. 1995, 196, 3267.
(40) Odian, G. G. Principles of Polymerization, 3rd ed.; J ohn Wiley
& Sons: New York, 1991.
(21) (a) Kuznetsov, A. N.; Wasserman, A. M.; Volkov, A. U.; Korst,
N. N. Chem. Phys. Lett. 1971, 12, 103. (b) Kovarskii, A. L.;
Wasserman, A. M.; Buchachenko, A. L. J . Magn. Reson. 1972,
7, 225. (c) Kivelson, D. J . Chem. Phys. 1960, 33, 1094. (d)
Stone, T. J .; Buckman, T.; Nordio, P. L.; McConnell, H. M.
Proc. Natl. Acad. Sci. U.S.A. 1965, 54, 1010.
(22) Kothe, T.; Marque, S.; Martschke, R.; Popov, M.; Fischer H.
J . Chem. Soc., Perkin Trans 2. 1998, 1553.
(23) Moffat, K. A.; Hamer, G. K.; Georges, M. K. Macromolecules
1999, 32, 1004.
(24) Fukuda, T.; Terauchi, T.; Goto, A.; Ohno, K.; Tsujii, Y.;
Miyamoto, T.; Kobatake, S.; Yamada, B. Macromolecules
1996, 29, 6396.
(25) Huang, W. Synthesis of mono-, bi-, and multinitroxide
radicals and application to the controlled radical polymeri-
(41) Hui, A. W.; Hamielec, A. E. J . Appl. Polym. Sci. 1972, 16,
749.
(42) (a) Fischer, H. Macromolecules 1997, 30, 5666. (b) Fischer,
H. J . Polym. Sci., Part A, Polym. Chem. 1999, 37, 1885. (c)
Souaille, M.; Fischer, H. Macromolecules 2000, 33, 7378.
(43) Lutz, J . F.; Lacroix-Desmazes, P.; Boutevin, B. Macromol.
Rapid Commun. 2001, 22, 189.
MA0116856