Polar, steric, and stabilization effects in alkoxyamines C-ON bond homolysis: A multiparameter analysis

Article abstract of DOI:10.1021/ma050004u

We present measurements of the rate constants (k_d) of the C-ON bond cleavage in new alkoxyamine models containing the N-(2-methyl-2-propyl)-N-diethylphosphono-2,2-dimethylpropyl)-N-oxyl (SG1) moiety. The homolysis rate constants of 2,2,6,6-tetramethylpiperidinyl-1-oxyl (TEMPO)-and SG1-based alkoxyamines are analyzed in terms of polar inductive/field (σ_U), steric (σv), and radical stabilization (σ_RS) contributions of the leaving alkyl radicals, using a multiparameter equation, i.e., log(k_d/k _d,0) = ρ_Uσ_U + δv + ρ _RSσ_RS. The rate constants increase with increasing electron withdrawing, steric, and stabilization demands of the leaving alkyl radicals. Good correlations are found for TEMPO (log(k_d/k _d,0) = 13.6σ_U + 6.6v + 13.9σ_RS) and SG1 (log(k_d/k_d,0) = 19.5σ_U + 7.0v + 15.3σ_RS) derivatives, highlighting the polar sensitivity of the leaving alkyl radical to the nitroxyl moiety. Such correlations should facilitate the design of new alkoxyamines as initiators/regulators and help to improve the tuning of NMP experiments.

Full text of DOI:10.1021/ma050004u

2638
Macromolecules 2005, 38, 2638-2650
Polar, Steric, and Stabilization Effects in Alkoxyamines C-ON Bond
Homolysis: A Multiparameter Analysis
Denis Bertin, Didier Gigmes, Sylvain R. A. Marque,* and Paul Tordo
UMR 6517 Chimie, Biologie et Radicaux Libres, case 542, Universite´ de Provence, Avenue Escadrille
Normandie-Niemen, 13397 Marseille Cedex 20, France
Received January 3, 2005; Revised Manuscript Received January 12, 2005
ABSTRACT: We present measurements of the rate constants (kd) of the C-ON bond cleavage in new
alkoxyamine models containing the N-(2-methyl-2-propyl)-N-(1-diethylphosphono-2,2-dimethylpropyl)-
N-oxyl (SG1) moiety. The homolysis rate constants of 2,2,6,6-tetramethylpiperidinyl-1-oxyl (TEMPO)-
and SG1-based alkoxyamines are analyzed in terms of polar inductive/field (σU), steric (υ), and radi-
cal stabilization (σRS) contributions of the leaving alkyl radicals, using a multiparameter equation, i.e.,
log(kd/kd,0) ) FUσU + δυ + FRSσRS. The rate constants increase with increasing electron withdrawing,
steric, and stabilization demands of the leaving alkyl radicals. Good correlations are found for TEMPO
(log(kd/kd,0) ) 13.6σU + 6.6υ + 13.9σRS) and SG1 (log(kd/kd,0) ) 19.5σU + 7.0υ + 15.3σRS) derivatives,
highlighting the polar sensitivity of the leaving alkyl radical to the nitroxyl moiety. Such correlations
should facilitate the design of new alkoxyamines as initiators/regulators and help to improve the tuning
of NMP experiments.
Scheme 1
Introduction
In their seminal works, Rizzardo et al.1 and later
Georges et al.2 have demonstrated that polymers with
polydispersities well below the theoretical limit of 1.5
for classical radical polymerization can be prepared by
nitroxide-mediated radical polymerization (NMP). Since
then, many groups3 have pushed this concept further
to the preparation of polymers with well-defined archi-
tectures, molecular weights, and compositions. These
polymerizations are governed by the persistent radical
effect (PRE).4-8 The simplest mechanism involves the
reversible dissociation of a dormant nitroxyl end-capped
polymer chain with n g 0 monomer units into a
Scheme 2. Radical Cyclization Reaction Using the
PRE
•
transient carbon-centered radical Rn and a persistent
nitroxyl radical (reactions 1 and 2 in Scheme 1), the
propagation of the carbon-centered radical (reaction 3,
Scheme 1), and its irreversible termination (reaction 4,
Scheme 1).6
In the course of time a quasi-equilibrium of the
reversible dissociation step (reactions 1 and 2, Scheme
1) is reached. This is unusual because there is a large
excess of persistent species and because the two con-
centrations are weakly time dependent. If this quasi-
equilibrium is rapidly established in comparison to the
monomer conversion, all of the chains grow uniformly
from n ) 0 on. A kinetic treatment9 with chain-length-
independent rate constants10 has shown that the po-
lymerization time decreases with increasing equilibrium
constants K ) kd/kc of the reversible dissociation of the
nitroxyl end-capped polymer alkoxyamine. That is, it
decreases with increasing values of kd. The polydisper-
sity index decreases with increasing conversion and
becomes small for large products kdkc. However, kd must
not exceed a critical value for which the controlling PRE
breaks down, and the optimum values of kd and kc
depend on the propagation and termination rate con-
stants (reactions 3 and 4, Scheme 1). In general, kd g
10-3 s-1 is desirable,11 and it would be helpful if kd could
be reasonably predicted on the basis of alkoxyamine and
nitroxyl radical structures to avoid unnecessary syn-
thetic and polymerization works.
Moreover, very recently, the first purposely done
experiments to use the PRE and alkoxyamines in
radical chemistry have been presented by Studer and
Ciufolini et al. (Scheme 2).12 Consequently, knowledge
of kd is also useful for the synthetic planning of nitroxyl-
mediated tin free radical chemistry.13,14
For selected cases, the enthalpic and steric factors
that influence kd have been addressed before.15-23 Stu-
dies on 2,2,6,6-tetramethylpiperidinyl-1-oxyl (TEMPO)-
alkoxyamines have recently shown a relationship be-
tween Ea or log kd and bond dissociation energy (BDE-
(C-ON)24,25 or BDE(C-H)13,26) of the leaving alkyl
radical, suggesting that stabilization of the leaving alkyl
radical is a major factor. Such a relationship also
seemed valid for other alkoxyamines based on N-(2-
* To whom correspondence should be addressed: Fax 33-4-
91-28-87-58; e-mail marque@srepir1.univ-mrs.fr.
10.1021/ma050004u CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/26/2005

Macromolecules, Vol. 38, No. 7, 2005
Alkoxyamines C-ON Bond Homolysis 2639
Scheme 3
cedure,46 stored at -20 °C, and had a purity higher than 98%.
Chromatography columns were performed using Merck silica
gel 60.
methyl-2-propyl)-N-(1-diethylphosphono-2,2-dimethyl-
propyl)-N-oxyl (SG1) and di-tert-butylnitroxyl (DBNO)
moieties.26 However, it gives only a rough estimation
of the value of kd and is strongly dependent on the
accuracy of the value of BDE(C-H). Indeed, other
effects such as steric and polar effects in the ground
state can bias the values of BDE (C-H).27-33 Further-
more, we have recently shown that the correlation log
kd vs BDE(C-H) for SG1 derivatives is more compli-
cated than expected from the well-known TEMPO
derivative case.34 To overcome these difficulties, we
present hereafter an analysis following a multiparam-
eter approach35 where log kd is parametrized with the
molecular descriptors corresponding to the polar induc-
tive/field36-41 (σU), steric42,43 (υ), and the stabilization44
(σRS) effects of the leaving alkyl radicals in the TEMPO-
and SG1-alkoxyamine series. Scheme 3 shows the
nitroxyl and alkyl moieties of the alkoxyamines.
General Procedure for the Synthesis of Alkoxyamines
2r and 2s. Under an inert atmosphere, a solution of SG1 (5
mmol, 1.47 g) and alkyl bromide (10 mmol, 1.67 g of 2-bro-
mobutyric acid or 2-bromo-2-methylpropionic acid) in benzene
(8 mL) was transferred to a mixture of CuBr (10 mmol, 1.43
g), N,N,N′,N′,N′′-pentamethyldiethylenetriamine (20 mmol,
3.47 g), and Cu(0) (5 mmol, 0.64 g) in benzene (8 mL). After 3
h stirring at room temperature, 30 mL of water was poured
into the mixture which was washed with diethyl ether (3 ×
20 mL). The aqueous layer was saturated with NH4Cl and then
extracted with diethyl ether (3 × 20 mL). The combined
organic layers were dried over anhydrous MgSO4. Removal of
the solvent under reduced pressure afforded 2r (1.34 g, 70%)
as a mixture of two diastereoisomers and 2s (1.43 g, 75%). The
diastereoisomers of 2r were further separated by silica gel
column chromatography (ethyl acetate).
N-(2-Methylpropyl)-N-(1-diethylphosphono-2,2-di-
methylpropyl)-O-(1-carboxylpropyl)hydroxylamine (2r).
Anal. Calcd for C17H36NO6P (381.44): C 53.53; H 9.51; N 3.67.
Found: C 53.54; H 9.66; N 3.91. Isomer A: 31P NMR (CDCl3):
Unfortunately, not all derivatives were available for
the two nitroxyl fragments 1 and 2 (see Table 1). Alkyl
radicals a, b, d, e, g, h, j, o, q, s, and t should
be reasonable models for the propagating radical of
R-methylstyrene, styrene, methacrylates, acrylates, meth-
acrylonitrile, acrylonitrile, butadiene, ethylene, iso-
prene, methacrylic acid, and acrylic acid, respectively.
3
27.0. 1H NMR (CDCl3): 0.88 (t, JH-H ) 14.0 Hz, 3H,
CHCH2CH3), 1.09 (s, 9H, C(CH3)), 1.12 (s, 9H, C(CH3)), 1.31
3
3
(t, JH-H ) 12.0 Hz, 3H, OCH2CH3), 1.29 (t, JH-H ) 12.0 Hz,
3H, OCH2CH3), 1.6-2.0 (m, 1H, CHCH2CH3), 2.23-2.36 (m,
1H, CHCH2CH3), 3.35 (d, 2JH-P ) 21.0 Hz, 1H, CHP), 3.9-4.4
(m, 5H, CH2CH3 and CHCH2). 13C NMR (CDCl3): 174.0
Experimental Section
1
(COOH), 86.8 (CH-COOH), 70.8 (d, JC-P ) 210 Hz, C-P),
All reactions were carried out in oven-dried glassware under
an argon atmosphere. Benzene (98%), copper bromide (98%),
copper powder (99%), N,N,N′,N′,N′′-pentamethyldiethylenetri-
amine (99%), and 2-bromobutyric acid (97%) were used as
received from Aldrich. The 2-bromo-2-methylpropionic acid
(98%) was purchased from Fluka and used as received.
TEMPO (98%, Lancaster) was sublimed, and SG1 (85% purity,
Arkema group) was used as received. The solvent tert-butyl-
benzene (t-BuPh, Aldrich) was purified by standard proce-
dures.45 NMR experiments were performed in CDCl3 as solvent
on a 300 Avance Bruker spectrometer (1H 300 MHz, 13C 75.48
MHz, and 31P 121.59 MHz) in “Centre Regional de RMN” in
Marseille. Chemical shifts were given with TMS as internal
reference for 1H NMR, CDCl3 (internal reference) for 13C NMR,
and H3PO4 85% (external reference) for 31P NMR. Elemental
analyses were done in “Service Commun de Micro Analyse
Universite´ d’Aix-Marseille 3”. Alkoxyamines 2r and 2s were
synthesized following a modified21,23,25,34 Matyjaszewski pro-
2
62.9 (C-N), 60.69 (d, JC-P ) 12.0 Hz, CH2-O-P), 60.2 (d,
2JC-P ) 10.6 Hz, CH2-O-P), 35.65 (d, JC-P ) 7.5 Hz,
2
3
P-CHCMe3), 30.0 (d, JC-P ) 8.5 Hz, CHC(CH3)3), 27.8
3
(C(CH3)), 16.42 (d, JC-P ) 8.7 Hz, P-OCH2CH3), 16.21 (d,
3JC-P ) 10.6 Hz, P-OCH2CH3), 9.3 (CH2CH3). Isomer B: 31P
1
3
NMR (CDCl3): 24.3. H NMR (CDCl3): 0.95 (t, JH-H ) 14.0
Hz, 3H, CHCH2CH3), 1.12 (s, 9H, C(CH3)), 1.14 (s, 9H, C(CH3)),
1.26 (t, 3JH-H ) 13.0 Hz, 3H, OCH2CH3), 1.28 (t, 3JH-H ) 13.0
Hz, 3H, OCH2CH3), 1.6-2.0 (m, 1H, CHCH2CH3), 2.23-2.36
(m, 1H, CHCH2CH3), 3.31 (d, 2JH-P ) 21.0 Hz, 1H, CHP), 3.9-
4.4 (m, 5H, CH2CH3 and CHCH2). 13C NMR (CDCl3): 173.3
1
(COOH), 87, 0 (CH-COOH), 68.1 (d, JC-P ) 206 Hz, C-P),
2
62.8 (C-N), 62.52 (d, JC-P ) 19.7 Hz, CH2-O-P), 62.16 (d,
2JC-P ) 9.3 Hz, CH2-O-P), 35.65 (d, JC-P ) 7.5 Hz,
2
3
P-CHCMe3), 30.4 (d, JC-P ) 8.7 Hz, CHC(CH3)3), 27.8
3
(C(CH3)), 16.42 (d, JC-P ) 8.7 Hz, P-OCH2CH3), 16.21 (d,
3JC-P ) 10.6 Hz, P-OCH2CH3), 9.7 (CH2CH3).

2640 Bertin et al.
Macromolecules, Vol. 38, No. 7, 2005
Table 1. Experimental Arrhenius Parameters (A and Ea), Rate Constants kd at 120 °C, and Corrected Activation Energies
Ecaorr for Alkoxyamines Based on the Nitroxyl Fragments 1 and 2 (Structures Shown in Scheme 3)
alkoxyamines
runs
T (°C)a
Ab (1014 s-1
)
Eac (kJ mol-1
)
kd,393d (s-1
)
Ecaorr e (kJ mol-1
)
ref
1a
1b
1c
1d
1e
1f
1g
1h
1j
9
70-92
2.0
2.5
(2.4)
1.8
1.0
(2.4)
6.8
8.9
(2.4)
(2.4)
115.7
133.0
145.5
119.8
139.0
161.5
118.4
137.9
126.4
139.2
79.4
165.0
145.8
192.3
124.5
134.4
128.4
130.8
149.1
121.9
125.2
127.4
136.0
131.9
130.0
126.4
169.2
162.3
139.7
128.5
112.3
132.9p
130.7q
8.5 × 10-2
5.2 × 10-4
1.1 × 10-5
2.2 × 10-2
3.4 × 10-5
8.1 × 10-8
0.13
116.2
132.9
26
26
26
26
26
26
26
13
26
26
24
17
2
90-150
131-150
65-121
90-151
151
16f
11f,g
2f
120.7
141.8
15
7
3
2
51-103
100-152
101-131
109-130
114.9
134.3
3.4 × 10-4
3.8 × 10-4
7.5 × 10-5
6.7 × 103
2.8 × 10-8
1.0 × 10-5
1k
1l
1p
1q
1u
2b
2c
2e
1
150
(2.4)
(2.4)
26, 24
3
130-151
503-533
60-137
110-131
80-132
90-129
130-150
100
13
2
6.6 × 10-12
5.5 × 10-3
3.3 × 10-4
3.0 × 10-3
1.0 × 10-3
3.6 × 10-6
1.5 × 10-2
5.5 × 10-3
2.8 × 10-3
2.0 × 10-4
7.1 × 10-4
1.3 × 10-3
3.8 × 10-3
7.8 × 10-9
6.4 × 10-8
6.5 × 10-5
2.0 × 10-3
0.28
24
20h
3
1.9
(2.4)
3.5
125.2
127.2
26
26
12i,j
4j,k
3
26, 70, 71
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
2f
2h
25
34, 70
1l
1l
2m
3
100
110
2i
120-140
100-130
90-120
100-137
150
34
2k
2m
2n
2o
2p
2q
2r
2s
2t
3
3
34
34
6h
1n
1n
3n
3o
3
34
25
150
25
120-140
112-124
60-81
100-120
25
this work
this work
70
4
5.3 × 10-4
1.0 × 10-3
a T ( 1 °C. b Statistical errors smaller than a factor 2. Value in brackets is the average of all experimentally accessible frequency
factors. c Statistical errors between 2 and 3 kJ mol-1 d Values calculated with parameters from columns 4 and 5. e Rescaled values of Ea
.
using a mean frequency factor (see text) when it was necessary; i.e., A different from 2.4 × 1014 s-1
.
f R ) t-Bu. g Values may be biaised
due to a side reaction; see ref 80. h Both diastereoisomers show same rate constant. i Isomers RS/SR, 31P NMR (C6D6), δ ) 22.3 ppm.j R
) Me. k Isomers RR/SS, 31P NMR (C6D6), δ ) 23.0 ppm. l Both isomers were not purified separately. Rate constants were estimated from
a plot of eq 1 showing two slopes. m Only one isomer could be measured. 31P NMR (C6D6) δ ) 21.8 ppm. n As kd is slow, same side reaction
as in ref 80 could occur. o Only one diastereoisomer (δ ) 24.3 ppm) was available pure. p δ ) 25.6 ppm in C6D6/t-BuPh. q δ ) 22.9 ppm
in C6D6/t-BuPh.
N-(2-Methylpropyl)-N-(1-diethylphosphono-2,2-di-
methylpropyl)-O-(2-carboxylprop-2-yl)hydroxylamine
(2s). Anal. Calcd for C17H36NO6P (381.44): C, 53.53; H, 9.51;
N, 3.67; Found: C, 53.57; H, 9.28; N, 3.77. 31P NMR (CDCl3):
26.2. 1H NMR (CDCl3): 1.15 (s, 9H, C(CH3)), 1.24 (s, 9H,
lifetime considerably. Therefore, kd cannot be taken
directly from its decay under normal reaction condi-
tions.4,7,8,26 Instead, conditions have to be chosen such
that the transient radicals are rapidly and completely
converted into other unreactive species (reaction 4)
before reaction 2 (Scheme 1) can occur. In our earlier
works,7,26 we showed that kd can be measured conve-
niently by quantitative CW-ESR spectroscopy from the
appearance of the nitroxyl radical signal using a suitable
alkyl radical scavenger such as galvinoxyl, TMIO-15ND12
(perdeuterated 2,2,10,10-tetramethylisoindolin-[15N]-N-
oxyl), TEMPO, and air (oxygen). Other authors have
employed nitroxyl radicals,18 styrene,15-23,47 good hy-
drogen donors,24,48 diiodine or nitroso compounds,49 and
oxygen.20,49,50 The experimental cleavage rate constants
kd were measured using either the plateau method (eq
1) or the initial slope (eq 2) of the time evolution of the
doubly integrated ESR signal of the growing nitroxyl
radical in the presence of oxygen as alkyl radical
scavenger.7,10,13,21,25,26
C(CH3)), 1.33 (t, 3JH-H ) 6 Hz, 3H, CH2CH3), 1.36 (t, 3JH-H
)
6 Hz, 3H, CH2CH3), 1.61 (s, 3H, CH3), 1.78 (s, 3H, CH3), 3.42
(d, 2JH-P ) 27 Hz, 1H, CHP), 3.9-4.4 (m, 4H, CH2). 13C NMR
(CDCl3): 176.5 (COOH), 84.5 (C-COOH), 70.8 (d, 1JC-P ) 136
2
Hz, C-P), 63.0 (C-N), 62.06 (d, JC-P ) 6.6 Hz, CH2-O-P),
2
2
59.94 (d, JC-P ) 7.5 Hz, CH2-O-P), 35.74 (d, JC-P
)
3
6.0 Hz, CMe3), 29.5 (d, JC-P ) 6.0 Hz, P-CHC(CH3)3), 27.4
(C(CH3)3), 27.9 (CCH3), 24.1 (CCH3), 16.15 (d, 3JC-P ) 6.0 Hz,
3
P-OCH2CH3), 15.77 (d, JC-P ) 7.5 Hz, P-OCH2(CH3)).
ESR Kinetic Experiments. ESR kinetic experiments were
performed on a CW-ESR spectrometer (Bruker EMX) in
conditions previously described,7,10,13,21,26 using air (oxygen) as
an alkyl radical scavenger and samples freshly prepared from
pure compounds.
Results
Measurements of the Rate Constant of the C-ON
Bond Homolysis. Here, in the absence of monomers,
we assumed that low molecular weight leaving alkyl
radicals mimic polymerizing systems with similar sub-
stituted propagating alkyl radicals. Since there is no
propagation in the present case and n ) 0, only reac-
tions 1, 2, and 4 (Scheme 1) should occur. Re-formation
of the alkoxyamine by reaction 2 increases its apparent
[nitroxide]t - [nitroxide]∞
ln
) -kdt
(1)
(2)
(
)
[nitroxide]∞
[nitroxide]t
[nitroxide]∞
) kdt

Macromolecules, Vol. 38, No. 7, 2005
Alkoxyamines C-ON Bond Homolysis 2641
Table 2. Values of the Universal Electrical Hammett Constant σU, the Steric Constant υ, the Corrected (RSEcorr
and Uncorrected (RSE) Radical Stabilization Energy, the Radical Stabilization Constants σRS and σ•, and
Bond Dissociation Energy of the C-H Bond BDE(C-H) of the Alkyl Fragments a-u (Structures Shown
in Scheme 3)
)
a
d
alkyl fragment
σU
υb
RSEc (kJ mol-1
)
RSEcorr d (kJ mol-1
)
σRS
σ• e
BDE(C-H)f (kJ mol-1
)
a
b
c
d
e
f
g
h
i
0.05
0.07
1.28g,h
0.86g,k
0.70
-35.1
-35.1
-35.1
-11.7
-11.7
-11.7
-14.2
-14.2
-14.2
-52.7
-52.7
-62.6hh
-27.9ll
-26.8oo
0.0
-53.1
-49.3
-44.7
-29.7
-25.9
-21.3
-32.2
-28.4
-23.8
-66.9
-62.3
-78.8
-37.5
-41.0
-9.6
-14.2
-18.0
-26.5
-30.3
-26.5
0.0
0.36
0.34
0.31
0.20
0.18
0.15
0.22
0.19
0.16
0.46
0.43
0.54
0.25
0.28
0.07
0.10
0.12
0.18
0.21
0.18
0.00
2.46i
2.11l
1.73
352.8 (6.3)j
357.0 (6.3)j
370.0 (6.3)m
379.0 (15.0)r
385.0 (15.0)r
406.3 (10.5)r
362.6 (8.4)z
375.8 (9.6)z
396.3 (8.8)z
354.0 (6.3)ff
368.8 (8.8)m
321.9 (6.3)jj
364.5 (8.4)j
385.8 (5.9)j
422.6 (1.7)qq
399.6 (4.0)m
404.0 (1.7)m
394.5tt
0.03
0.07n
0.09s
0.15
1.43g,o,p
1.00g,p,t
0.80g,p,v
1.20g,x
0.79g,aa
0.58g,cc
0.86g,dd
0.69
1.16q
0.81u
0.43w
1.15y
1.10bb
0.72
0.14n
0.17s
0.20
j
k
l
0.03s
0.02
2.43ee
2.05
0.07s
0.12kk
0.09s
-0.01kk
0.00
1.50d,gg
0.72
-
ii
m
n
o
p
q
r
1.5mm
0.84pp
0.42
0.82g,nn
0.73
0.87
1.24
0.0
0.80
-0.01
0.09s
0.07n
0.09s
-0.01
0.0
1.15
1.02g,rr
1.24g,uu
0.83g,xx
0.52
-12.3ss
-12.3ss
-12.3ss
0.0
0.78tt
1.13vv
0.78yy
0.00
s
388.5 (12.1)ww
394.5 (12.1)zz
438.5 (0.4)m
t
u
a Values given in ref 58 unless otherwise mentioned. b Values given in refs 42 and 43 unless otherwise mentioned. c Values given in ref
55 unless otherwise mentioned. One assumed that primary alkyl radicals exhibit same RSE as secondary and tertiary. d See text. e Values
given in ref 57 unless otherwise mentioned. f Errors are given in parentheses. g Estimated from eq 13. h υ1 ) υ(t-Bu) and υ2
)
υ(Ph) ) 0.57. i Value of σ•(PhCH2) incremented of 0.73; see ref 57. j Reference 76. k υ1 ) υ(i-Pr) ) 0.76 and υ2 ) υ(Ph) ) 0.57. l Value of
σ•(PhCH2) incremented of 0.38; see ref 57. m Reference 64. n Estimated from eq 12. o υ1 ) υ(t-Bu) and υ2 ) υ(CO2Alk) ) 0.9. p In the
case of TEMPO derivatives, the type of R group (Me or t-Bu) does not modify the values of kd unlike for SG1 derivatives; see refs 81
and 82. q Value of σ•(MeOOCH2) estimated in footnote w incremented of 0.73; see ref 57. r For d and e groups, see ref 77. For f group,
see ref 83. s Estimated from eq 11. t υ1 ) υ(i-Pr) ) 0.76 and υ2 ) υ(CO2Alk) ) 0.9. u Value of σ•(MeOOCH2) estimated in footnote w
incremented of 0.38; see ref 57. v υ1 ) υ(Et) 0.52 and υ2 ) υ(CO2Alk) ) 0.9. w Value estimated using eq 32 and aH,â(MeOOCCMeH•) )
24.68 G from refs 55 and 84. x υ1 ) υ(t-Bu) and υ2 ) υ(CN) ) 0.40. y Value of σ•(NCCH2) incremented of 0.73; see ref 57. z Reference
78. aa υ1 ) υ(i-Pr) ) 0.76 and υ2 ) υ(CN) ) 0.40. bb Value of σ•(NCCH2) incremented of 0.38; see ref 57. cc υ1 ) υ(Et) ) 0.52 and υ2
)
υ(CN) ) 0.40. dd υ1 ) υ(i-Pr) ) 0.76 and υ2 ) υ(CHdCH2) ) 0.57. ee Value of σ•(H2CdCHCH2) incremented of 0.38; see ref 57. ff Ref-
erence 26. gg A mean value between υ(Ph) and υ(C6H11) was assumed. hh Estimated from eq 3 in ref 55 and aH,R ) 13.56 G from
ref 72. ii Cannot be estimated using eq 32. jj Reference 24. kk Estimated from eq 10. ll Estimated with eq 3 in ref 55 and aH,R
)
11.88 G of MeCtCCHMe• from ref 76. mm Value estimated using eq 32 and aH,â(MeCtCCMeH•) ) 19.16 G from ref 76. nn υ1 ) υ(i-Pr) )
•
0.76 and υ2 ) υ(Ac) ) 0.50. oo Estimated with eq 3 and aH,R ) 19.7 G of MeCOCH2 from ref 55. pp Value of σ•(H3CCOCH2) ) 0.94
incremented of 0.38; see ref 57. qq BDE(C-H) of C2H6 was used for C6H14; see ref 76. rr υ1 ) υ(s-Bu) ) 1.02 and υ2 ) υ(CO2H) )
•
0.50. ss Estimated from eq 3 in ref 55 and aH,â ) 21.6 G of HOOCMe2 and aH,R ) 20.1 from HOOCHMe• from refs 74 and 75. Both
values were averaged. tt One assumed that the presence of a methyl group in â position did not affect the values of σ• and BDE(C-H) to
a large extent. uu υ1 ) υ(t-Bu) and υ2 ) υ(CO2H) ) 0.50. vv Value of σ•(HOOCCH2) ) 0.40 incremented of 0.73; see ref 57.
ww BDE(HOOCMe2-H) ) BDE(MeOOCMe2-H) + 9.5 kJ mol-1. Increment given in ref 79. xx υ1 ) υ(i-Pr) ) 0.76 and υ2 ) υ(CO2H) )
0.50. yy Value of σ•(HOOCCH2) ) 0.40 incremented of 0.38; see ref 57. zz BDE(HOOCMeH-H) ) BDE(MeOOCMeH-H) + 9.5 kJ mol-1
.
Increment given in ref 79.
The experiments were carried out in t-BuPh and SG1
used as the standard nitroxyl radical. We showed earlier
that the frequency factor A for the C-ON bond homoly-
sis for 30 alkoxyamines did not vary much with the
alkoxyamine structure but lay between 1013 and 1015
depending on the substituent carried by the leaving
alkyl radical) when Ea was plotted against BDE(C-H)
(eqs 3a and 3b).
Ea kJ mol-1 ) -133.0 (250) kJ mol-1 + 0.72 (7) ×
BDE(C-H) kJ mol-1 R2 ) 0.97 (3a)
s-1 with an averaged value A of 2.4 × 1014 s-1 7,13,21,25,26,34
.
For 2s and 2r, the rate constants were determined only
at a few temperatures. Thus, the activation energies Ea
were estimated directly from the rate constants kd using
the averaged frequency factor A. The estimated Ea
values given in Table 1 correspond to the average of
individual values over the temperature range. Indi-
vidual values vary by less than 2 kJ mol-1 from the
average values listed in Table 1. The number of experi-
ments, temperature ranges, frequency factors A, ac-
tivation energies Ea, cleavage rate constants at 120 °C
Ea kJ mol-1 ) -137.0 (780) kJ mol-1 + 0.69 (13) ×
BDE(C-H) kJ mol-1 R2 ) 0.77 (3b)
The regression coefficient for the polar line is not good
(R2 ) 0.77) due to a scattering of the data because the
differences in Ea for the two isomers of 2e, 2h, and 2t
were not taken into account and due to the strongly
deviating value of Ea of 2f. Such deviation might be
caused either by the large error on BDE(C-H) of
H-CH2COOMe (see Table 2) or by a biased value of Ea
due to side reactions. Alkoxyamine 2r is not included
in the regression because the statistic parameters are
poorer (R2 ) 0.61). That is certainly due to some steric
effect of the ethyl group of the r group. However, the
value of Ea for 2r lies close to the line corresponding to
kd,393, and corrected activation energies Ecorr for al-
a
koxyamines based on the nitroxyl fragments 1 and 2
are listed in Table 1.
Bond Dissociation Energy BDE(C-H). We have
recently shown34 that the values of Ea for the homoly-
sis of the CO-N bond of SG1-alkoxyamines were
divided into two families (one nonpolar and one polar

2642 Bertin et al.
Macromolecules, Vol. 38, No. 7, 2005
Figure 1. Activation energies Ea for the C-ON bond homoly-
sis of alkoxyamines based on the nitroxyl fragment 2 vs bond
dissociation energy BDE(C-H) of the corresponding hydro-
carbon (structures shown in Scheme 3): (9) nonpolar (primary
and secondary) leaving alkyl radicals; (b) polar (primary and
secondary); and (2) tertiary leaving alkyl radical (polar and
nonpolar).
Figure 2. Activation energies Ea for C-ON bond homolysis
in alkoxyamines based on the nitroxyl fragment 1 vs radical
stabilization energy (RSE) of the leaving alkyl radicals: (9)
tertiary, (b) secondary, and (2) primary leaving alkyl radicals
(structures shown in Scheme 3).
the polar leaving alkyl radicals (Figure 1). In contrast,
2s lies far from the two lines (polar and nonpolar radi-
cals), which suggests a possible influence of the steric
effect. Indeed, such a result could explain the position
of 2q on the line for polar leaving alkyl radicals (vide
infra).
Radical Stabilization Energy. Some years ago,
Ru¨chardt et al.51-55 defined a radical stabilization
energy (RSE), which was assumed free of any of the
effects observed with BDE(C-H). Therefore, we plotted
Ea vs RSE for TEMPO (Figure 2) and SG1 (Figure 3)
derivatives. Assuming the same RSE for primary,
secondary, and tertiary leaving alkyl radicals (Table
2), three lines (eqs 4-6) were drawn for TEMPO
derivatives.
Figure 3. Activation energies Ea for C-ON bond homolysis
in alkoxyamines based on the nitroxyl fragment 2 vs radical
stabilization energy (RSE) of the leaving alkyl radicals: (b)
secondary and (2) primary leaving alkyl radicals (structures
shown in Scheme 3).
Ecaorr (kJ mol-1) )
167.0 ((3.4) + 0.55 ((0.09) × RSE (4)
R2 ) 0.97, σ ) 2.7, N ) 3
factor to take into account. Further comments will be
developed in the Discussion section.
Ecaorr (kJ mol-1) )
Ecaorr (kJ mol-1) )
159.0 ((10.3) + 0.99 ((0.28) × RSE (5)
R2 ) 0.76, σ ) 15.4, N ) 6
156.3 ((7.3) + 0.62 ((0.25) × RSE (7)
R2 ) 0.60, s ) 10.0, N ) 6
Ecaorr (kJ mol-1) )
135.5 ((9.4) + 0.73 ((0.48) × RSE (6)
Ecaorr (kJ mol-1) )
R2 ) 0.54, σ ) 12.0, N ) 4
142.5 ((5.8) + 0.75 ((0.33) × RSE (8)
R2 ) 0.39, s ) 9.4, N ) 10
The straight line for the primary (eq 4) alkyl radicals
shows a high linear correlation coefficient (R2 ) 0.97),
which suggests a major influence of RSE on kd. The
straight lines for the secondary (eq 5) and tertiary (eq
6) exhibit a poor R2 (Figure 2). This scattering is
certainly caused by some steric effect arising from the
methyl group(s) on the carbon of the C-ON bond or by
the peculiar conformation of the two six-membered rings
(p and l) for the secondary alkyl radical.
Although two straight lines (primary and secondary
alkyl radicals) were drawn for SG1 derivatives, Figure
3 shows the scattered plots of Ea vs RSE which are
underlined by small R2 values (eq 7 for primary and eq
8 for secondary alkyl radicals). It is clear that the
stabilization of the leaving alkyl radical is not the only
Taft-Ingold Approach. To overcome the several
difficulties27 inherent in the use of BDE(C-H) as pre-
dictive tool and to gain more insight into the effects in-
volved in the C-ON bond homolysis of the alkoxyamines,
we investigated a multiparameter approach.35 Indeed,
one of us has already used this approach to disentangle
the polar and steric effects of the nitroxyl moiety acting
on the rate constant of the homolysis of the alkoxyamine
C-ON bond.56 From the previous approaches (vide
supra) applied to the TEMPO and SG1 derivatives, it
was obvious that the polar, steric, and radical stabiliza-
tion effects of the leaving alkyl radicals influence kd and
are differently intertwined depending on the type of

Macromolecules, Vol. 38, No. 7, 2005
Alkoxyamines C-ON Bond Homolysis 2643
Scheme 4
and tertiary) of alkyl radical studied. However, these
values are larger than those deduced from the BDE-
(C-H) of the corresponding alkane.64 We supposed that
some ground-state effects biased them. Consequently,
we preferred to use the values given by Ru¨chardt et al.53
and reported them in eqs 14-16 to give the corrected
radical stabilization energy (RSEcorr, Table 2), RSE-
nitroxyl moiety. Thus, eq 9 should help to disentangle
these effects by applying polar inductive/field (σU),36-41
steric (υ),42,43 and radical stabilization (σRS or σ•)44,57
molecular descriptors to the leaving alkyl radicals.
(CH3 ) ) 0.0 kJ mol-1 being the reference.
•
RSEcorr(tertiary) ) RSE - 18.0 kJ mol-1 (14)
log kd ) log kd,0 + FRSσRS + FUσU + δυ
(9)
RSEcorr(secondary) ) RSE - 14.2 kJ mol-1
As the various substituents were directly bound to the
reactive center (Scheme 4), the value of σU of the
CR1R2R3 group was used (Table 2).58
The unknown values were given by eqs 10-12 for
primary, secondary, and tertiary alkyl groups, respec-
tively.59
(15)
RSEcorr(primary) ) RSE - 9.6 kJ mol-1 (16)
To normalize the RSEcorr values, we used the formation
•
enthalpy of the methyl radical (∆Hf(CH3 ) ) -146.3 kJ
mol-1) as it should contain a minimum of effects due to
the saturated ground state of the molecule.64 The values
of σRS are given by eq 17.
σU,R CH ) 0.416σU,R - 0.0103
(10)
(11)
(12)
1
2
1
σU,R R CH ) 0.297
σ
U,R + 0.00482
∑
1
2
RSEcorr
σRS
)
(17)
•
∆Hf(CH3 )
σU,R R R C ) 0.248
σU,R + 0.00398
∑
1
2 3
For the steric demand, the constant of Charton42,43
υ
The larger σRS is, the more stabilized the alkyl radical
is. Before applying the σRS values to the multiparametric
regressions, it had been checked that they were not
correlated to σU (R2 ) 0.01) and υ (R2 ) 0.02).
(Table 2) was preferred to the Taft60 constants Es
because more groups were available, and the values
were expected to be free of inductive and resonance
effects.42,43 The missing values were given by eq 13.61
A previous work13 showed that a partial positive
charge was present on the carbon of the polarized
C-ON bond. Consequently, σR, σR+, and σR- (molecular
descriptors for delocalization effect) were tested.36,58 In
υ ) 0.866υ1 + 0.436υ2 - 0.0455
(13)
Values of υ1 and υ2 were given by Charton.42,43,61
Furthermore, Charton showed that the υ2 values for the
ester group varies between 0.50 and 1.39 depending on
the value of the dihedral angle θ between the COOR
group and the reactive center.61 Hence, for 1e (R ) Me)
and for 2e (R ) Me), X-rays showed θ values of 60° and
of 30°, respectively;62 thus, a mean value of θ ) 45 °
(i.e., υ2 ) 1.0) could be assumed for all derivatives in
both series.61 Intuitively, the l group (Scheme 3) looks
more sterically demanding than the cyclohexyl (υ )
0.87)42 but less than the phenyl group (υ ) 2.15).42 We
thus chose a value υ ) 1.5 to represent the size of the
cyclohexadienyl group.61
The main problems arose with the radical stabiliza-
tion molecular descriptor (σRS). In fact, the Hammett
constants σU and σR (delocalization effect) are not the
most suitable to take into account the radical stabiliza-
tion.28 A lot of work had already been done in that field
but mainly on the benzylic-type compounds.63 To the
best of our knowledge, only Afanas’ev57 developed in the
1970s a general scale for alkyl radical stabilization (vide
infra). However, over the past two decades, Ru¨chardt’s
group51-55 has developed a RSE scale for a large series
of alkyl radicals. Those RSE values were assumed to
be free of the ground-state effects involved in the
homolysis of C-X bonds (X being any atom or group).51-55
Moreover, the RSE values include the effects due to the
electron-withdrawing, electron-donating, and resonance
capacities of the various groups bound to the radical
center. The RSE values are the same whatever the type
of alkyl radicals studied and are given in kJ mol-1
(Table 2) and are not dimensionless like σU and υ.55 The
difference in the y-intercept of eqs 4-6 may be assumed
as the stabilization due to the type (primary, secondary,
the absence of one of the major parameters i.e., σU, σRS,
and υ, R2 and F-test values were very poor, and a four-
parameter regression did not significantly improve the
correlation. Therefore, for the moment, no experimental
results justify the use of the resonance Hammett
constants.
In earlier works,13,26 plotting Ea(C-ON) vs BDE-
(C-H) showed that, for TEMPO derivatives, kd was
mainly influenced by the stabilization energy and also
to a smaller extent by the size of the leaving alkyl
radical. In contrast, plotting Ea(C-ON) vs BDE(C-H)
for SG1 derivatives (Figure 1) emphasized the influence
of the polarity of the leaving alkyl radical, besides the
two precited effects.34 Hence, all the possible correlation
combinations with σRS, σU, and υ were tested and
collected in Tables 3 and 4. For the TEMPO derivative
series, eqs 18-20 suggest that radical stabilization is
the major factorsbut the statistical data (R2 ) 0.59) are
worse than with BDE(C-H) as expectedsand also
that the effect of the size of the leaving alkyl radical is
not negligible (R2 ) 0.49) whereas the polarity seems
unimportant (R2 ) 0.10). The statistical data of eqs
18-20 are improved by eqs 21 and 23 (biparametric
regressions). They could be further improved by remov-
ing deviating data such as those corresponding to the
tertiary alkyl radicals or compounds 1b,c,j and k.
Equation 22 (Table 3) shows very good R2 and F-test
values, which confirms the influence of the size and
stabilization of the leaving alkyl radical. Equation 24
(three parameters) improves the statistical data of eq
22. It will thus be considered as the equation giving the
better description of the effects influencing C-ON bond
homolysis in alkoxyamines (Figure 4, vide infra). The
discrepancy observed with 1f is certainly due to the

2644 Bertin et al.
Macromolecules, Vol. 38, No. 7, 2005
Table 3. Correlations of log kd at 120 °C with Various Molecular Descriptors (Radical Stabilization Constant σRS
,
Universal Electrical Hammett Constant σU, and Steric Constant υ) for Alkoxyamines Based on the Nitroxyl Fragment 1
eq
log kd,0
FRS
FU
δ
Na
sb
R2 c
td
Fe
18f
19
-8.4 ((1.3)
-5.0 ((1.4)
-11.3 ((2.2)
-9.5 ((1.3)
17.7 ((4.3)
14
14
13g
14
2.4
3.5
2.2
2.2
0.59
0.10
0.49
0.68
99.87
19.3 ((16.6)
18.4 ((10.3)
77.00
20
7.2 ((2.2)
99.20
21h
17.6 ((3.9)
13.8 ((2.2)
-
-
99.97i
91.80j
99.99i
99.99m
72.00j
99.77m
99.99i
99.45j
99.99m
99.82k
99.99n
97.60p
99.99q
22l
23o
24
-14.3 ((1.1)
-13.2 ((2.3)
-14.8 ((0.7)
7.0 ((1.1)
8.7 ((2.2)
6.6 ((0.7)
14
14
14
1.2
2.4
0.8
0.91
0.63
0.96
12.8 ((11.3)
13.6 ((3.2)
13.9 ((0.9)
a Number of data. b Standard deviation. c Square of the coefficient of the linear regression. d Student t-test given in percent. e F-test
given in percent. f Tertiary alkyl radicals were out of the general trend but included in the regression. g 1l was not included because it
was far from the straight line. h Tertiary alkyl radicals were slightly deviating from the regression line. i t-test on the value of FRS j t-test
.
on the value of FU. k F-test value is 12. l 1g,h were slightly deviating. m t-test on the value of δ. n F-test value is 56. o 1b,c,j,k were slightly
deviating. p F-test value is 6. q F-test value is 80.
Table 4. Correlations of log kd at 120 °C with Various Molecular Descriptors (Radical Stabilization Constant σRS
,
Universal Electrical Hammett Constant σU, and Steric Constant υ) for Alkoxyamines Based on the Nitroxyl Fragment 2
eq
log kd,0
FRS
FU
δ
Na
sb
R2 c
td
Fe
25
26
27
28i
-11.7 ((0.8)
-3.6 ((0.2)
-5.6 ((0.9)
-6.9 ((1.0)
49.6 ((4.6)
14f
14g
15h
19
0.7
0.4
0.6
1.4
0.89
0.57
0.49
0.49
99.99
9.0 ((2.3)
99.81
3.6 ((1.0)
99.66
11.1 ((3.9)
12.6 ((4.2)
13.8 ((5.1)
99.00j
98.40k
99.10j
92.40m
99.13k
95.55m
99.99j
99.99k
99.99m
99.52l
98.13n
98.17o
99.99p
29
30
31
-9.4 ((2.2)
-9.1 ((2.2)
-14.3 ((1.3)
4.1 ((2.2)
4.9 ((2.2)
7.0 ((1.1)
19
19
19
1.5
1.5
0.8
0.39
0.39
0.85
17.7 ((5.9)
19.5 ((3.0)
15.3 ((2.2)
a Number of data. b Standard deviation. c Square of the coefficient of the linear regression. d Student t-test given in percent. e F-test
given in percent. f 2b,c,k,m, and n were not included in the regression because they were far from the trend; R2 ) 0.25. g 2f,i,o,p, and
s were not included in the regression because they were far from the trend; R2 ) 0.21. h 2f,o,p, and q were not included in the regression
because they were far from the trend; R2 ) 0.06. i Tertiary alkyl radical were deviating from the regression line. j t-test on FRS
.
k t-test on
FU. l F-test value is 8. m t-test on δ. n F-test value is 5. o F-test value is 5. p F-test value is 29.
tested and are collected in Table 4. As expected from
Figure 1 and eqs 7 and 8, log kd can be correlated to
σRS, σU, and υ neither individually (eqs 25-27) nor
dually (eqs 28-30), even by removing some deviating
data whereas the R2 value was lower than 0.2 when the
19 data were used. In contrast, the correlation with
three parameters (eq 31 and Figure 5) exhibits signifi-
cantly improved R2 and F-test values. Equation 31
confirms the assumptions done from Figure 1 about the
effects influencing the C-ON bond homolysis. Here also
2f lies clearly far from the regression line. That behavior
is certainly due to the reasons mentioned above for the
TEMPO series. The scattering of the SG1 derivatives
looks worse than that of TEMPO derivatives because
the multiple parameter regression cannot account for
the different kd of the diastereoisomers.
Figure 4. Log kd at 120 °C for the C-ON bond homolysis in
alkoxyamines based on the nitroxyl fragment 1 vs linear
combination of molecular descriptor (radical stabilization
constant σRS, universal electrical Hammett constant σU, and
steric constant υ) for eq 24 (structures shown in Scheme 3).
Radical Hammett Constant σ•. As mentioned above,
only Afanas’ev57 has developed a general σ• scale based
either on kinetic measurements or on the hyperfine
coupling constant of the â hydrogen of the alkyl radical
MeXHC• (eq 32).
position of the ester group; i.e., such a group can adopt
any position, from the least to the most sterically
demanding position.60,61
σ• ) 5.16 - 0.192aHâ(G)
(32)
For the SG1 derivatives, the plot Ea(C-ON) vs BDE-
(C-H) (Figure 1) suggests clearly that the polar, radical
stabilization and steric effects due to the leaving alkyl
radical are strongly intertwined, each one acting sig-
nificantly. As for TEMPO derivatives, all of the possible
linear regression combinations with σRS, σU, and υ were
That σ• scale is highly correlated to our σRS (R2 ) 0.80),
and therefore the statistical data listed in Tables 5 and
6 and Tables 3 and 4 are similar. Nevertheless, the
correlations are clearly better when σRS values are used
instead σ• values due to (i) σRS (see eqs 14-17) is more

Macromolecules, Vol. 38, No. 7, 2005
Alkoxyamines C-ON Bond Homolysis 2645
and secondary alkyl radicals were drawn (Figure 3, but
with poor statistical tests), the scattered plots exemplify
unambiguously that the radical stabilization does not
have a major influence on the C-ON bond cleavage, as
expected from Figure 1. Indeed, for both series, a
multiple parameter approach35 (eqs 21-24 and 28-31)
turned out to be very powerful because the values of kd
were correlated with three parameters: σU (Hammett
constants for the universal electrical effect, i.e., localized
electrical or polar inductive/field effect),36-41,58,59 σRS
(Hammett constants for the radical stabilization effect),
and υ (constants for the steric effect).42,43,61 For the
TEMPO-alkoxyamine series, good correlations were
obtained with one (σRS, eq 18) or two (σRS and υ, eq 22)
parameters, but the three-parameter linear regression
(eq 24) gave even better results that can be compared
to the results obtained with the SG1-alkoxyamine
series.
Figure 5. Log kd at 120 °C for the C-ON bond homolysis in
alkoxyamines based on the nitroxyl fragments 2 vs linear
combination of molecular descriptor (radical stabilization
constant σRS, universal electrical Hammett constant σU, and
steric constant υ) for eq 31 (structures shown in Scheme 3).
Using eqs 41 and 42,65 with Xi the value of the
parameter X for the ith data, Xh the arithmetic mean of
the X parameter, n the number of data, ν the number
of degree of freedom, RX the weighting coefficient, and
CX the coefficient of the X parameter, it was possible to
estimate the weight of the various effects in each series.
general than σ•, i.e., almost any σRS of alkyl radical can
be estimated; (ii) σRS values are based on RSE values
and only correlated to hyperfine coupling constants of
R or â hydrogen of alkyl-centered radical;55 (iii) such
values should be mainly free of ground-state effects;51-55
(iv) some discrepancies (see Table 2) were observed
between the values of σRS and σ•, e.g., σ•(CMe2CN) <
σ•(t-Bu) ....
n
(X - Xh)2
∑
i
i)1
RX )
(41)
(42)
x
ν
Discussion
RXCX
R C
We34 showed that, for the SG1-alkoxyamines (Figure
1 and eqs 3a,b), the leaving alkyl radical carrying a
polar group bound to the radical center cleaves faster
contrary to the TEMPO derivatives for which no influ-
ence of the polarity was observed, e.g., kd(2h) ≈ kd(2b).
As expected, the values of Ea for 2t (t being a polar
group) lie close to the straight line corresponding to
polar leaving alkyl radicals. The large gap between Ea
for 2s and the straight line ascribed to the polar leaving
alkyl radicals confirms our previous assumption34 about
the behavior of 2q. That is, besides a polar effect acting
on kd, there is also a steric effect involved. In Figure 1,
the gaps between Ea of 2q and Ea of 2s and the lines
representing eqs 7 and 8-18 and 20 kJ/mol, respec-
tivelysare quite similar, which suggests that the steric
effect acts similarly in both series. Then, when a “polar”
nitroxyl radical is used, one needs to draw four different
correlations (polar and nonpolar primary plus second-
ary, and the same for tertiary alkyl fragments) to take
into account the major effects involved in the C-ON
bond homolysis. Consequently, plotting Ea vs BDE-
(C-H) is not the best tool to predict values of kd easily
and accurately or to determine and analyze the effects
influencing the C-ON bond cleavage. In fact, BDE-
(C-H) is composed of various effects involved in the
stability of the molecule in the ground state.27-33
For a better insight into the factors influencing the
C-ON bond homolysis, we plotted Ea vs RSE (radical
stabilization energy)51-55 for TEMPO (Figure 2) and
SG1 (Figure 3) derivatives. For TEMPO-alkoxyamines,
the three straight lines in Figure 2 show that the radical
stabilization is the major effect involved in the C-ON
cleavage. The scattered plot for the secondary and
tertiary alkyl radicals suggests that the size of the
leaving alkyl radical has an influence. In contrast, with
SG1 derivatives, although two parallel lines for primary
% X ) 100 ×
∑
Xi Xi
i
Entry 1 (Table 7) reveals that, in TEMPO series,
radical stabilization (44%) and steric (40%) demands are
the main effects involved in the C-ON bond homolysis.
Indeed, the polar field/inductive effect (16%) is weak but
not negligible. Those results are in good agreement with
what was observed in Figure 2 and earlier works.13,25,26
In contrast, in SG1 series (entry 2, Table 7), the absence
of major effects of the leaving alkyl radicalssradical
stabilization effect: 34%; steric effect: 31%; and the
polar (universal electrical) effect: 35%sconfirms the
remarks made about Figure 3 (no main influence of
radical stabilization) and about Figure 1 (four correla-
tions to explain the results). Although SG1 is bulkier
than TEMPO, its size has a smaller influence than
TEMPO’s (35% for TEMPO against 31% for SG1).
Moreover, the radical stabilization effect is lower in the
SG1-alkoxyamine series (34%) than in the TEMPO-
alkoxyamine series (44%). Conversely, the influence of
the electrical effect (σU) is 2 times higher in the SG1-
alkoxyamine series (35%) than in TEMPO-alkoxyamine
series (16%), which suggests that the leaving alkyl
radical is highly sensitive to the polarity of the nitroxyl
moiety.
All of these results are accounted for by changes in
the transition state (TS), ground state (GS), and final
state (FS) and are sketched in Figure 6 for nonpolar
leaving alkyl radicals. For both the nitroxyl and alkyl
moieties, an increase of kd with the increasing steric
demand is a feature of a destabilization in GS. Other-
wise, in the TS increasing steric demand involves a
smaller ∆Sq; thus, a larger ∆Gq and consequently kd is
smaller. With respect to the Hammond principle,66 one

2646 Bertin et al.
Macromolecules, Vol. 38, No. 7, 2005
Table 5. Correlations of log kd at 120 °C with Various Molecular Descriptors (Radical Stabilization Constant σ•,
Universal Electrical Hammett Constant σU, and Steric Constant υ) for Alkoxyamines Based on the Nitroxyl Fragment 1
eq
log kd,0
F•
FU
δ
Na
sb
R2 c
td
Fe
33
34
-7.9 ((1.2)
-9.7 ((1.1)
2.7 ((0.8)
2.9 ((0.6)
13
13
2.2
1.7
0.49
0.72
99.30
99.90f
98.32g
99.77f
99.63i
99.99f
99.89g
99.97i
23.0 ((8.0)
99.84h
99.86j
99.99k
35
36
-13.1 ((1.6)
-13.8 ((0.9)
2.3 ((0.6)
2.5 ((0.3)
6.0 ((1.6)
5.2 ((0.9)
13
13
1.5
0.8
0.79
0.94
18.9 ((4.0)
a Number of data. b Standard deviation. c Square of the coefficient of the linear regression. d Student t-test given in percent. e F-test
given in percent. f t-test on F•. g t-test on FU. h F-test value is 13. j F-test value is 19. i t-test on δ. k F-test value is 47.
Table 6. Correlations of log kd at 120 °C with Various Molecular Descriptors (Radical Stabilization Constant σ•,
Universal Electrical Hammett Constant σU, and Steric Constant υ) for Alkoxyamines Based on the Nitroxyl Fragment 2
eq
log kd,0
F•
FU
δ
Na
sb
R2 c
td
Fe
37
38
-5.0. ((1.0)
-6.9 ((1.0)
1.5 ((0.8)
2.0 ((0.7)
19
19
1.7
1.4
0.17
0.47
91.70
16.0 ((5.3)
98.67f
99.20g
93.27f
76.40i
99.34f
99.99g
99.82h
99.37h
88.67j
99.84k
39
40
-7.5 ((2.3)
1.6 ((0.8)
2.3 ((0.5)
2.9 ((2.3)
5.9 ((1.6)
19
19
1.7
1.0
0.24
0.73
-12.7 ((1.7)
21.9 ((4.2)
a Number of data. b Standard deviation. c Square of the coefficient of the linear regression. d Student t-test given in percent. e F-test
given in percent. f t-test on F•. g t-test on FU. h F-test value is 7. i t-test on δ. j F-test value is 3. k F-test value is 13.
Table 7. Weighting Coefficients for Eqs 24 and 31
ΣErelax covers the contribution of Ei for all bonds in XRZ
except for the cleaved XR-Z bond, and it increments
the value of Ei positively or negatively. Thus, Nau67 has
established that the PGSE between a molecule reference
HR-Z and a homologue XR-Z is roughly described by
eq 44.
entry
%σRS
%σU
%υ
1
2
44
34
16
35
40
31
can assume a late transition state, i.e., resembling FS.
Therefore, any stabilization of both nitroxyl and alkyl
radicals (FS) stabilizes TS and thus decreases Ea
(increasing kd). Oddly, an increase of kd with the
increasing polar effect was not expected because the
electron-withdrawing groups destabilize both the ni-
troxyl and alkyl moieties in TS. Indeed, one of us has
recently showed that electron-withdrawing groups on
the nitroxyl radical destabilize FS (the form B is not
stabilized by electron-withdrawing groups, Scheme 5)
and therefore TS; i.e., the more electron-withdrawing
the groups are, the smaller the values of kd are.13,56
Moreover, previous works suggest the presence of a
partial positive charge on the carbon of the C-ON bond
in TS.13 One expects, a priori, a destabilization of TS
with the increasing polarity of the alkyl moiety. Against
these expectations, kd is enhanced by electron-with-
drawing groups on both moieties.
All of these effects caused by the polarity of both the
nitroxyl and alkyl moieties are due to the polar ground-
state effect (PGSE) described by Nau.67 Moreover, in a
previous work, one of the authors has already pointed
out the influence of PGSE on the C-ON bond homolysis
in para-substituted 1b derivatives.13 Such effect is very
well-known for aromatic compounds,33 recently29-33
extended to any molecules, and it should account for
our observations. Briefly, Pauling68 has shown that BDE
includes a polar term (ionic resonance energy Ei) which
depends on the difference of electronegativity ø for the
atoms or groups of the cleaved bond. From these
grounds, Nau67 has shown that BDE of the XR-Z
molecule is described by eq 43.
PGSE ≡ 23[(øHR - øZ)2 - (øXR - øZ)2] - ∆ΣErelax
(44)
A consequence of that approach is that the more
mismatched the ø values are, the stronger the bond is.
Assuming that eq 44 is also true when Z is a group of
atoms and HR or XR is not merely a benzylic-type
group, and setting, thus, O-TEMP (1) or O-SG (2) for
Z and HR for nonpolar groups (NPG) and XR for polar
groups (PG), eqs 45a and 45c describe the PGSE for the
TEMPO and SG1 families, respectively.
PGSE1 ) 23[(øNPG - øOTEMP)2 - (øPG - øOTEMP)2] -
1
∆ΣErelax (45a)
∆ΣErelax1 ) 23(øPG - øNPG)(øNPG - øPG - 2øTEMP
(45b)
)
PGSE2 ) 23[(øNPG - øOSG)2 - (øPG - øOSG)2] -
2
∆ΣErelax (45c)
∆ΣErelax2 ) 23(øPG - øNPG)(øNPG - øPG - 2øSG) (45d)
When eqs 45a and 45c are combined to give eq 46a, it
is obvious that the larger the difference of ø between 1
and 2 is, the stronger the ∆PGSE is and the more
sensitive to the polar effect the C-ON bond is.
∆PGSE ) 46(øOSG - øOTEMP)(øNPG - øPG) -
∆∆ΣErelax (46a)
BDE(XR-Z) ) 1/2BDE(XR-RX) + 1/2BDE(Z-Z) +
23(øXR - ø )2 +
E
(43)
∆∆ΣErelax ) 46(øPG - øNPG)(øSG - øTEMP
)
(46b)
∑
Z
relax

Macromolecules, Vol. 38, No. 7, 2005
Alkoxyamines C-ON Bond Homolysis 2647
Figure 6. Influence of the steric (SD) and radical stabilization (RSD) demand on the ground state (GS), transition state (TS),
and final state (FS) of alkoxyamines based on the nitroxyl fragments 1 and 2 and carrying nonpolar alkyl moieties. Bold and
dotted arrows are for alkyl and nitroxyl moieties, respectively.
Scheme 5
fragment family (see footnotes of Table 8). Therefore,
in first approximation, the nitroxyl moieties belonging
to the nonpolar family should exhibit σU values (polar
effect) smaller than 0.10, and those belonging to the
polar family σU values larger than 0.10. Assuming that
the coefficients (FU, FRS, and δ) determined for the non-
polar or polar nitroxyl fragment family can be applied
respectively to any nitroxyl fragments belonging to one
of the two families, thus the difference in influence of
the nonpolar or polar nitroxyl fragment is contained in
the constant term log kd,0, i.e., the y-intercept which
accounts for all effects due to the substituents of the
nitroxyl fragment. Hence, the constant terms log kd,0
for a few current nitroxyl fragments are given in Table
8 considering they belong either to the nonpolar or to
the polar nitroxyl fragment family. Because of their σU
values, the nitroxyl radicals 1, 3, and 4 belong to the
nonpolar nitroxyl fragment family, but estimates of
their constant terms log kd,0 (values in italics) are also
given in Table 8 considering the coefficients (FU, FRS, and
δ) for the polar nitroxyl fragment. Similarly, because
of their σU values, nitroxyl radicals 2, 5, and 6 belong
to the polar nitroxyl fragment family, but estimates of
their constant terms log kd,0 (values in italics) are also
given in Table 8 considering the coefficients (FU, FRS, and
δ) for the nonpolar nitroxyl fragment family. Then, it
is possible to estimate kd values for alkoxyamines
containing various nitroxyl fragments.
The EPR nitrogen hyperfine coupling constants and
the multiparametric approach of the effect of the ni-
troxyl moiety on the value of kd support the view of
radical 2 being more polar than radical 1. Then, it is
not surprising to observe a stronger PGSE in the SG1-
alkoxyamine series than in the TEMPO-alkoxyamine
series. It is worthwhile to point out that the difference
in energy between polar and nonpolar groups due to the
PGSE is roughly of 15 kJ/mol, which is probably the
highest value observed for such effect. In general, for
bond homolysis such a value is assumed around 5-10
kJ/mol.67,69 The PGSE for the alkoxyamine 2 family is
depicted in Scheme 6.67 One can assume a polar Cδ+
-
δ-ON bond, and thus any electron-withdrawing group
(EWG) destabilizes the ground state (Scheme 6a), and
reversely for electron-donating groups (EDG) (Scheme
6b).
Hence, to predict a value of kd from the structure of
the alkoxyamine, the nitroxyl moiety has to be divided
in two families, one polar and one nonpolar. In a recent
work,56 it has been shown that the polarity of the
nitroxyl moiety is estimated with eq 47.
6
Conclusion
σ )
σ
(47)
∑
U
U,i
We have shown that the multiparameter approach
provides a better and deeper insight into the effects
(polar, steric, and radical stabilization demands) ruling
the C-ON bond homolysis in alkoxyamines than any
other approaches (BDE(C-H), RSE, or monoparametric
correlations). Hence, we were able to disentangle the
various intertwined effects (polar, steric, and radical
stabilization effects), and we pointed out that the
presence and the strength of each effect is dependent
on the structure of the nitroxyl radical counterpart. It
was shown that the polar effect depends strongly on the
structure of the nitroxyl fragment; that is, for the
i)1
The difference (≈9.0 kJ mol-1) of Ea between 3e and
3b is close to the difference observed between 1e and
1b (8.9 kJ mol-1) rather than the one between 2e and
2b (4.0-6.3 kJ mol-1), which implies 3 (Scheme 7)
belongs to the nonpolar nitroxyl family (see footnotes
of Table 8). On the other hand, the difference25 (6.5 kJ
mol-1) of Ea between 5e and 5b is close to the difference
observed between 2e and 2b (4.0-6.3 kJ mol-1) rather
than the one between 1e and 1b (≈9.0 kJ mol-1), which
implies 5 (Scheme 7) belongs to the polar nitroxyl

2648 Bertin et al.
Macromolecules, Vol. 38, No. 7, 2005
Scheme 6
Scheme 7
Table 8. Values of log kd,0 for Some Useful Nitroxyl
constants kd with only a short glance to the structure
of the alkoxyamine considered. Such system of equa-
tions should quickly turn out to be indispensable to
design new alkoxyamines and to tune NMP experi-
ments.
Fragmentsa
1
3b
4c
5d
6e
2
f
σU
log kd,0
nonpolar
polar
-0,06
-0.06
0.08
0.10
0.33
0.28
-14.9g -13.2g -14.0 -13.8g -13.5 -13.3g
-16.0g -14.5g -15.2 -15.0g -14.7 -14.4g
Acknowledgment. We thank Arkema Co., CNRS,
and University of Provence for financial support. S.M.
is deeply grateful to Professors M. Charton and C.
Ru¨chardt for helpful and fruitful discussions and com-
ments.
a Estimated from eq 24 (nonpolar family) and eq 31 (polar
family). Values in italics are for the estimation in the opposite
family; see text. b DBNO is di-tert-butyl-N-oxyl radical. 3b: A )
2.2 × 1014 s-1, Ea ) 121.8 kJ mol-1. 3e: A ) 1.2 × 1014 s-1, Ea )
128.2 kJ mol-1. See ref 26. With A ) 2.4 × 1014 s-1, Ea(3b) ) 122.1
kJ mol-1 and Ea(3e) ) 130.4 kJ mol-1
.
c 4b: A ) 2.4 × 1014 s-1
,
References and Notes
Ea ) 128.0 kJ mol-1. See ref 56. d TIPNO is 2,2,5,5-tetramethyl-
4-phenyl-3-azahexane-3-oxyl radical. 5b: A ) 5.6 × 1014 s-1, Ea
) 129.6 kJ mol-1, see ref 26. 5e: A ) 2.4 × 1014 s-1, Ea ) 133.4
kJ mol-1, see ref 25. With A ) 2.4 × 1014 s-1, Ea(5b) ) 126.9 kJ
(1) US Patent 4,581,429; Eur. Pat. Appl. 135280. Solomon, D.
H.; Rizzardo, E.; Cacioli, P. Chem. Abstr. 1985, 102, 221335q.
(2) Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer,
G. K. Macromolecules 1993, 26, 2987-2988.
mol-1
.
e 6b: A ) 2.4 × 1014 s-1, Ea ) 124.2 kJ mol-1. See ref 56.
(3) Hawker, C. J. Acc. Chem. Res. 1997, 30, 373-382. Hawker,
C. J.; Bosman, A. W.; Harth, E. Chem. Rev. 2001, 101, 3661-
3668 and references therein. Tunca, U.; Ozyurek, Z.; Erdogan,
T.; Hizal, G. J. Polym. Sci., Part A: Polym. Chem. 2004, 42,
4228-4236. Gillies, E. R.; Fre´chet, J. M. J. J. Am. Chem. Soc.
2002, 124, 14137-14146. Schierholz, K.; Givehchi, M.; Fabre,
P.; Nallet, F.; Papon, E.; Guerret, O.; Gnanou, Y. Macromol-
ecules 2003, 36, 5995-5999.
f Estimated with eq 47. g Averaged value of the b and e derivatives.
weakly polar nitroxyl fragment, the influence of the
polarity of the leaving alkyl group on the homolysis is
weak, whereas for the polar nitroxyl fragment, the
influence of the polarity of the leaving alkyl group on
the homolysis is strong. Moreover, combining eqs 24 and
31 and eq 8 (log(kd/s-1) ) 3.07σI - 0.88Es - 5.88) of
Marque56 gives a powerful tool to predict accurate rate
(4) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925-3927.
(5) Daikh, B. E.; Finke, R. G. J. Am. Chem. Soc. 1992, 114, 2938-
2943.

Macromolecules, Vol. 38, No. 7, 2005
Alkoxyamines C-ON Bond Homolysis 2649
(6) Fischer, H. Macromolecules 1997, 303, 55666-5672.
(7) Kothe, T.; Marque, S.; Martschke, R.; Popov, M.; Fischer, H.
J. Chem. Soc., Perkin Trans. 2 1998, 1553-1559.
(8) Fischer, H. Chem. Rev. 2001, 101, 3581-3610 and references
therein.
(9) Fischer, H. J. Polym. Sci., Part A: Polym. Chem. 1999, 37,
1885-1901. Souaille, M.; Fischer, H. Macromolecules 2000,
33, 7378-7394. Souaille, M.; Fischer, H. Macromolecules
2001, 34, 2830-2838.
(10) For kd values: Bertin, D.; Chauvin, F.; Marque, S.; Tordo, P.
Macromolecules 2002, 35, 3790-3791.
(42) Charton, M. Top. Curr. Chem. 1983, 114, 57-91.
(43) Charton, M. In The Chemistry of Dienes and Polyenes;
Rappoport, Z., Ed.; John Wiley & Sons Ltd.: 1997; Vol. 1, pp
683-732.
(44) σRS is a new Hammett constant defined later in that paper.
(45) Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals; 5th ed.; Butterworth Heinemann: Amsterdam,
2003.
(46) Matyjaszewski, K.; Woodworth, B. E.; Zhang, X.; Gaynor, S.;
Metzner, Z. Macromolecules 1998, 31, 5955-5957.
(47) Stipa, P.; Greci, L.; Carloni, P.; Damiani, E. Polym. Degrad.
Stab. 1997, 55, 323-327.
(48) Korolev, G. V.; Berezin, M. P.; Bakova, G. M.; Kochneva, I.
S. J. Polym. Sci., Ser. B 2000, 42, 339-344.
(11) Fischer, H.; Souaille, M. Chimia 2001, 55, 109-113. Fischer,
H.; Souaille, M. Macromol. Symp. 2001, 174, 231-240.
(12) Studer, A. Angew. Chem., Int. Ed. 2000, 36, 1108-1111.
Leroi, C.; Fenet, B.; Couturier, J. L.; Guerret, O.; Ciufolini,
M. A. Org. Lett. 2003, 5, 1079-1081. Leroi, C.; Bertin, D.;
Dufils, P.-E.; Gigmes, D.; Marque, S.; Tordo, P.; Couturier,
J. L.; Guerret, O.; Ciufolini, M. A. Org. Lett. 2003, 26, 4943-
4945.
(49) Kovtun, G. A.; Aleksandrov, A. L.; Golubev, V. A. Bull. Acad.
Chim. USSR, Div. Chem. 1974, 10, 2115-2121.
(50) Howard, J. A.; Tait, J. C. J. Org. Chem. 1978, 43, 4279-
4283.
(51) Beckhaus, H.-D.; Ru¨chardt, C. Chem. Ber. 1977, 110, 878-
(13) Marque, S.; Fischer, H.; Baier, E.; Studer, A. J. Org. Chem.
2001, 66, 1146-1156.
895.
(52) Ru¨chardt, C. Top. Curr. Chem. 1980, 88, 1-32.
(53) Ru¨chardt, C.; Beckhaus, H.-D. Top. Curr. Chem. 1985, 1-22.
(54) Birkhofer, H.; Beckhaus, H.-D.; Ru¨chardt, C. In Substituent
Effects in Radical Chemistry; Viehe, H. G., Janousek, Z.,
Mere´niy, R., Eds.; NATO ASI Series C; Kluiwert Academy
Press: Dordrecht, The Netherlands, 1986; Vol. 189, pp 199-
218.
(14) Allen, A. D.; Cheng, B.; Fenwick, M. H.; Givehchi, B.; Henry-
Riyad, H.; Nikolaev, V. A.; Shikhova, E. A.; Tahmassebi,
D.; Tidwell, T. T.; Wang, S. J. Org. Chem. 2001, 66,
2611-2617. Allen, A. D.; Fenwick, M. H.; Henry-Riyad, H.;
Tidwell, T. T. J. Org. Chem. 2001, 66, 5759-5765. Henry-
Riyad, H.; Tidwell, T. T. Can. J. Chem. 2003, 81, 697-704.
Henry-Riyad, H.; Tidwell, T. T. J. Phys. Org. Chem. 2003,
16, 559-563.
(55) Brocks, J. J.; Beckhaus, H.-D.; Beckwith, A. L. J.; Ru¨chardt,
C. J. Org. Chem. 1998, 63, 1935-1943.
(15) Moad, G.; Rizzardo, E. Macromolecules 1995, 28, 8722-8728.
(16) Goto, A.; Tomoya, T.; Fukuda, T.; Miyamoto, T. Macromol.
Rapid Commun. 1997, 673-681.
(17) Goto, A.; Fukuda, T. Macromol. Rapid Commun. 1997, 683-
688.
(18) Skene, W. G.; Belt, S. T.; Connolly, T. J.; Hahn, P.; Scaiano,
J. C. Macromolecules 1998, 31, 9103-9105.
(56) Marque, S. J. Org. Chem. 2003, 68, 7582-7590.
(57) Afanas’ev, I. B. Int. J. Chem. Kinet. 1975, 7, 857-877.
(58) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119-251.
Charton, M. Prog. Phys. Org. Chem. 1987, 16, 287-315.
(59) Charton, M. Adv. Quant. Struct. Prop. Relat. 1996, 1, 171-
219.
(60) Kutter, E.; Hansch, C. J. Med. Chem. 1969, 12, 7-652. Fujita,
T.; Takayama, C.; Nakajima, M. J. Org. Chem. 1973, 38,
1623-1631. Unger, S. H.; Hansch, C. Prog. Phys. Org. Chem.
1976, 12, 91-118. MacPhee, J. A.; Panaye, A.; Dubois, J.-E.
Tetrahedron 1978, 34, 3553-3562. Fujita, T. Pure Appl.
Chem. 1978, 50, 987-994. Hansch, C.; Leo, A. Substituent
Constants for Correlation Analysis in Chemistry and Biology;
John Wiley & Sons: New York, 1979; pp 1-167. Dubois,
J.-E.; MacPhee, J. A.; Panaye, A. Tetrahedron 1980, 36, 919-
928. Fujita, T.; Iwamura, H. Top. Curr. Chem. 1983, 114,
119-157.
(19) Ohno, K.; Tsujii, Y.; Miyamoto, T.; Fukuda, T.; Goto, M.;
Kobayashi, K.; Akaike, T. Macromolecules 1998, 31, 1064-
1069.
(20) Bon, S. A. F.; Chambard, G.; German, A. L. Macromolecules
1999, 32, 8269-8276.
(21) Le Mercier, C.; Lutz, J.-F.; Marque, S.; Le Moigne, F.; Tordo,
P.; Lacroix-Desmazes, P.; Boutevin, B.; Couturier, J. L.;
Guerret, O.; Martschke, R.; Sobek, J.; Fischer, H. ACS Symp.
Ser. 2000, Chapter 8, 108-122.
(22) Goto, A.; Fukuda, T. Macromol. Chem. Phys. 2000, 201,
2138-2142.
(23) Benoit, D.; Grimaldi, S.; Robin, S.; Finet, J. P.; Tordo, P.;
Gnanou, Y. J. Am. Chem. Soc. 2000, 122, 5929-5939.
(24) Ciriano, M. V.; Korth, H.-G.; Van Scheppingen, W. B.; Mulder,
P. J. Am. Chem. Soc. 1999, 121, 6375-6381.
(25) Le Mercier, C.; Acerbis, S.; Bertin, D.; Chauvin, F.; Gigmes,
D.; Guerret, O.; Lansalot, M.; Marque, S.; Le Moigne, F.;
Fischer, H.; Tordo, P. Macromol. Symp. 2002, 182, 225-
247.
(26) Marque, S.; Le Mercier, C.; Tordo, P.; Fischer, H. Macromol-
ecules 2000, 33, 4403-4410.
(27) Ru¨chardt, C. Angew. Chem., Int. Ed. Engl. 1970, 9, 830-
843.
(28) Cherkasov, A. R.; Jonsson, M.; Galkin, V. I.; Cherkasov, R.
A. Russ. Chem. Rev. 2001, 70, 1-22 and references therein.
(29) Ingold, K. U.; Wright, J. S. J. Chem. Educ. 2000, 77, 1062-
1064.
(61) Charton, M. Stud. Org. Chem. 1991, 42, 629-692.
(62) Le Mercier, C.; Acerbis, S.; Gigmes, D.; Marque, S.; Bertin,
D.; Siri, D.; Tordo, P., to be published.
(63) Creary, X. J. Org. Chem. 1980, 45, 280-284. Dust, J. M.;
Arnold, D. R. J. Am. Chem. Soc. 1983, 105, 1221-1227, 6531.
Arnold, D. R.; Nicholas, A. M. D. P.; Snow, M. S. Can. J.
Chem. 1985, 63, 1150-1155. Arnold, D. R. In Substituent
Effects in Radical Chemistry; Viehe, H. G., Janousek, Z.,
Mere´niy, R., Eds.; Kluiwert Academy Press: Dordrecht, The
Netherlands, 1986; Vol. 189, pp 171-188. Creary, X. Ibid.,
pp 245-262. Jackson, A. R. Ibid., pp 325-328. Creary, X.;
Mehrsheikh-Mohammadi, M. E.; McDonald, S. J. Org. Chem.
1987, 52, 3254-3263. Nau, W. M.; Harrer, H. M.; Adam, W.
J. Am. Chem. Soc. 1994, 116, 10972-10982. Jackson, R. A.;
Sharifi, M. J. Chem. Soc., Perkin Trans. 2 1996, 775-778.
Jiang, X.-K. Acc. Chem. Res. 1997, 30, 283-289. Adam, W.;
Harrer, H. M.; Kita, F.; Korth, H.-G.; Nau, W. M. J. Org.
Chem. 1997, 62, 1419-1426. Creary, X.; Engel, P. S.;
Kavaluskas, N.; Pan, L.; Wolf, A. J. Org. Chem. 1999, 64,
5634-5643.
(64) Berkowitz, J.; Ellison, G. B.; Gutman, D. J. Phys. Chem. 1994,
98, 2744-2765.
(65) Shorter, J. In Correlation Analysis of Organic Reactivity; J.
Wiley & Sons: New York, 1982; pp 73-126.
(66) Smith, M. B.; March, J. In Advanced Organic Chemistry, 5th
ed.; John Wiley & Sons: New York, 2001; p 284.
(67) Nau, W. J. Phys. Org. Chem. 1997, 10, 445-455.
(68) Pauling, L. The Nature of the Chemical Bond; Cornell
University Press: Ithaca, NY, 1960.
(30) Zavitsas, A. A. J. Chem. Educ. 2001, 78, 417-419.
(31) Coote, M. L.; Pross, A.; Radom, L. Org. Lett. 2003, 5, 4689-
4692.
(32) Matsunaga, N.; Rogers, D. W.; Zavitsas, A. A. J. Org. Chem.
2003, 68, 3158-3172.
(33) Pratt, D. A.; DiLabio, G. A.; Mulder, P.; Ingold, K. U. Acc.
Chem. Res. 2004, 37, 334-340 and references cited herein.
(34) Bertin, D.; Gigmes, D.; Le Mercier, C.; Marque, S. R. A.;
Tordo, P. J. Org. Chem. 2004, 69, 4925-4930.
(35) Shorter, J. In Correlation Analysis of Organic Reactivity; J.
Wiley & Sons: New York, 1982; pp 9-25.
(36) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165-
195 and references therein.
(69) Pratt, D. A.; Wright, J. S.; Ingold, K. U. J. Am. Chem. Soc.
1999, 121, 4877-4882.
(37) Exner, O. J. Phys. Org. Chem. 1999, 12, 265-274.
(38) Galkin, V. I. J. Phys. Org. Chem. 1999, 12, 283-288.
(39) Charton, M. J. Phys. Org. Chem. 1999, 12, 275-282.
(40) Exner, O.; Charton, M.; Galkin, V. J. Phys. Org. Chem. 1999,
12, 289.
(70) Bertin, D.; Gigmes, D.; Marque, S.; Tordo, P. e-Polym. 2003,
paper 2.
(71) Anantchenko, G. S.; Souaille, M.; Le Mercier, C.; Tordo, P.;
Fischer, H. J. Polym. Sci., Part A: Polym. Chem. 2002, 4,
3264-3283.
(41) Charton, M.; Charton, B. I. J. Phys. Org. Chem. 2001, 14,
832-838.

2650 Bertin et al.
Macromolecules, Vol. 38, No. 7, 2005
(72) Berndt, A. Landolt-Bo¨rnstein; Fisher, H., Hellweg, K.-H.,
Eds.; Springer-Verlag: Berlin, 1977; Vol. II/9, Part b, p 452.
(73) Berndt, A. Landolt-Bo¨rnstein; Fisher, H., Hellweg, K.-H.,
Eds.; Springer-Verlag: Berlin, 1977; Vol. II/9, Part b, p 430.
(74) Fischer, H.; Paul, P. Landolt-Bo¨rnstein; Fisher, H., Hellweg,
K.-H., Eds.; Springer-Verlag: Berlin, 1977; Vol. II/9, Part b,
p 189.
(79) Henry, D. J.; Parkinson, C. J.; Mayer, P. M.; Radom, L. J.
Phys. Chem. A 2001, 105, 6750-6756.
(80) Ananchenko, G. S.; Fischer, H. J. Polym. Sci., Part A: Polym.
Chem. 2001, 39, 3604-3621.
(81) Ananchenko, G. S.; Marque, S.; Gigmes, D.; Bertin, D.; Tordo,
P. Org. Biomol. Chem. 2004, 2, 709-714.
(82) Bertin, D.; Gigmes, D.; Marque, S. A. R.; Maurin, R.; Tordo,
P. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 3504-
3515.
(83) Luo, Y.-R. Handbook of Bond Dissociation Energies in
Organic Compounds; CRC Press: Boca Raton, FL, 2003; p
66.
(84) Knu¨hl, B.; Marque, S.; Fischer, H. Helv. Chim. Acta 2001,
84, 2290-2299.
(75) Fischer, H.; Paul, P. Landolt-Bo¨rnstein; Fisher, H., Hellweg,
K.-H., Eds.; Springer-Verlag: Berlin, 1977; Vol. II/9, Part b,
p 93.
(76) McMillen, D. F.; Golden, D. M. Annu. Rev. Phys. Chem. 1982,
33, 493-532.
(77) Zytowski, T.; Fischer, H. J. Am. Chem. Soc. 1997, 119,
12869-12878.
(78) King, K. D.; Goddard, R. D. J. Phys. Chem. 1976, 80, 546-
552.
MA050004U