Steric and polar effects of the cyclic nitroxyl fragment on the C-ON bond homolysis rate constant

Article abstract of DOI:10.1021/ma0512612

Alkoxyamines and persistent nitroxyl radicals are important regulators of nitroxide mediated radical polymerization (NMP). Because polymerization times decrease with increasing rate constant of the homolysis of the C-ON bond between the polymer chain and the nitroxyl moiety, the factors influencing the cleavage rate constant are of considerable interest. Here, we present the measurements of the rate constants (k_d) of the C-ON bond cleavage in new cyclic alkoxyamine models. The homolysis rate constants of 9 new alkoxyamines and 33 others given by the literature are analyzed with regards to the contributions of the polar inductive/field (σ_I) effect, the steric (E _s) effect and the intramolecular hydrogen bonding (IHB) effect of the nitroxyl moieties, using the multiparameter equation established by Marque, i.e., log(k_d/k_d,0) = -3.07σ_I - 0.88E _s - 5.88. Cyclic steric constants r(R_i) for seven- and eight-membered rings are developed. Analysis of the results provides new insight on the importance of the conformation of the alkoxyamine on the values of k_d.

Full text of DOI:10.1021/ma0512612

9974
Macromolecules 2005, 38, 9974-9984
Steric and Polar Effects of the Cyclic Nitroxyl Fragment on the C-ON
Bond Homolysis Rate Constant
Hanns Fischer,†,‡ Andreas Kramer,# Sylvain R. A. Marque,§* and Peter Nesvadba*,#
Coating Effects Research, Ciba Specialty Chemicals Inc., Schwarzwaldallee 215, CH-4002, Basel,
Switzerland, University of Zurich, CH-8057, Zurich, Switzerland, and UMR 6517 Chimie,
Biologie et Radicaux Libres, case 542, Universite´ de Provence, Avenue Escadrille Normandie-Niemen,
13397 Marseille Cedex 20 France
Received June 16, 2005; Revised Manuscript Received September 20, 2005
ABSTRACT: Alkoxyamines and persistent nitroxyl radicals are important regulators of nitroxide mediated
radical polymerization (NMP). Because polymerization times decrease with increasing rate constant of
the homolysis of the C-ON bond between the polymer chain and the nitroxyl moiety, the factors
influencing the cleavage rate constant are of considerable interest. Here, we present the measurements
of the rate constants (kd) of the C-ON bond cleavage in new cyclic alkoxyamine models. The homolysis
rate constants of 9 new alkoxyamines and 33 others given by the literature are analyzed with regards to
the contributions of the polar inductive/field (σI) effect, the steric (Es) effect and the intramolecular
hydrogen bonding (IHB) effect of the nitroxyl moieties, using the multiparameter equation established
by Marque, i.e., log(kd/kd,0) ) -3.07σI - 0.88Es - 5.88. Cyclic steric constants r(Ri) for seven- and eight-
membered rings are developed. Analysis of the results provides new insight on the importance of the
conformation of the alkoxyamine on the values of kd.
Scheme 1
Introduction
Since the pioneering work of Rizzardo1 and Georges,2
nitroxide mediated free radical polymerizations (NMP)
have often been used for the synthesis of living polymers
and copolymers with low polydisperties, controlled mo-
lecular weights, and elaborate architectures.3,4 The ideal
mechanism is outlined in Scheme 1. Nitroxide-capped
chain molecules cleave into propagating alkyl and
persistent nitroxyl radicals (kd). The propagating radi-
cals grow by monomer addition (kp) and re-form longer
dormant chains by cross-coupling with the nitroxyl
radicals (kc). Simultaneously, the usual self-termination
of the propagating radicals into unreactive polymer
products takes place (kt). This removes the propagating
radicals and increases nitroxyl radical concentration in
time; the cross-coupling reaction therefore dominates
over self-termination. For long reaction times, a quasi-
equilibrium of the reversible cleavage is reached, char-
acterized by weakly time-dependent concentrations and
a large excess of the persistent over the propagating
radicals.5-9
Among other factors, the polar and steric effects of
the substituents of both the leaving alkyl23-25,27-36 and
the nitroxyl23-26,29,36-46 radicals affect the homolysis rate
constant kd. Nevertheless, although systematic studies
on the nitroxyl moiety are scarce,36,38,40,46 one of us46 has
recently developed a Taft-Ingold approach that sepa-
rates the polar and steric effects of the substituents on
the nitroxyl moiety in the sum (eq 1) of the universal
electrical Hammett constant47-51 σI and the Taft steric
constant52,53 Es.
log kd ) -3.07σI,n - 0.88Es,n - 5.88
(1)
For the well-controlled and living radical polymeri-
zation of a given monomer with propagation and ter-
mination constants kp and kt, the rate constants of the
reversible cleavage kd and kc must fall into proper
ranges.6,8-12 Several studies have shown that kd is a
crucial parameter because it varies within a broad
range, i.e., from 2.4 to 10-10 s-1, whereas the range for
kc is rather narrow (106-109 L mol-1 s-1).7,8,10-15
Furthermore, at a given time, the polydispersity index
is lower for a larger kd.8,10,11,13,15 Consequently, the
determination of kd has recently found considerable
attention for both polymeric16-22 and low molecular
weight model systems.23-45
We present hereafter the Arrhenius parameters and the
kd values for eight piperidine-squeleton-based (11, 25,
26, 32-34, 36, 36a) one PROXYL-squeleton-based (18),
and two diazepanone-based (20, 44) alkoxyamines (Table
1). Those values, along with the kd of 33 alkoxyamines
from the literature36-42,54 (Scheme 2), are analyzed with
eq 1 to determine the influence of the polar effect, the
ring size effect, the ring substitution and the ring sp2/
sp3 hybridization effects, the leveled steric effect, and
the intramolecular hydrogen bonding (IHB) effect.
Experimental Section
* Corresponding author. E-mail: peter.nesvadba@cibasc.com.
† Deceased February, 22, 2005.
The syntheses of alkoxyamines 11, 18 20, 25, 26, 32-34,
and 44 have recently been described.55-59
‡ University of Zurich.
Remarks. Due to the presence of several diastereoisomers,
§ Universite´ de Provence. Fax: 33-4-9-28-87-58. E-mail: marque@
srepir1.univ-mrs.fr.
1
the H and 13C NMR spectra are extremely complicated, and
# Ciba Specialty Chemicals Inc. E-mail (P.N.): peter.nesvadba@
cibascs.com.
comprehensive assignment of individual resonances was there-
fore not possible. Hence, only unambiguous signals were
10.1021/ma0512612 CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/03/2005

Macromolecules, Vol. 38, No. 24, 2005
C-ON Bond Homolysis Rate 9975
Table 1. Arrhenius Parameters (A, Ea), Rate Constants
(kd) at 120 °C for C-ON Bond Homolysis in
Scheme 2
Alkoxyamines 11, 18, 20, 25, 26, 32-34, 36, 36a, and 44,
and Nitrogen Hyperfine Coupling Constants aN of the
Corresponding Nitroxyl Radicals
a Statistical error smaller than a factor of 2. Values in brackets
are the average of all experimentally accessible frequency factors
given in refs 23, 24, 36, and 38. b Statistical error between 2 and
3 kJ/mol. c EPR nitrogen hyperfine coupling constant aN in tert-
butylbenzene. The corresponding nitroxyl radicals 18• 20•, 25•, 26•,
32•, 33•, 34•, 36•, and 44• are displayed in Scheme 6. d The
reestimated value of Ea with A ) 2.4 × 1014 s-1 is of 114 kJ/mol.
e The reestimated value of Ea is of 112 kJ/mol with A ) 2.4 × 1014
s-1
.
individually assigned. The identification of the new compounds
rests on these typical 1H and 13C NMR assignments, combined
with very good elemental analysis and gas chromatography
coupled with mass spectrometer (GC-MS) analysis.
General Data. 1H NMR (300 MHz) and 13C NMR (75.37
MHz) spectra were recorded in CDCl3 on a Bruker 300
spectrometer. IR spectra were taken on a Nicolet Magna-IR
750 spectrometer in KBr pill, MS spectra on a Finnigan SSQ
710 apparatus using isobutane for chemical ionization (CI).
Elemental analyses were performed on a Leco CHNS-932
apparatus. Gas chromatography was performed on Agilent
6850 Series chromatograph on a HP-1 methyl siloxane capil-
lary column (30 m × 320 µm × 0.25 µm) with He carrier gas
(3 mL/min) and FID detector, injector temperature 250 °C,
starting temperature 100 °C, and gradient 20°/min to 300 °C.
Materials. The following reagents and solvents were ac-
quired from Fluka and were used as received: acetanhydride
(g99%), tert-butyl hydroperoxide (70% in water), copper(I)
chloride (g97%), copper(I) bromide (g98%), 2-chloropropionic
acid methyl ester (g97%), 4-(dimethylamino)pyridine (g98%),
ethyl acetate (g99.5), ethylbenzene (g98%), hexane (g99.5%),
methyl tert-butyl ether (g99%), N,N,N′,N′′,N′′-pentamethyl-
diethylene-triamine (g98%), pentane (g99.5%), sodium boro-
hydride (g96%), tetrabutylammonium bromide (g98%), and
toluene (g99.7%). Peracetic acid (40% in acetic acid, Peraclean)
was from Degussa. All other chemicals were from Fluka and
were g98% pure. Silica gel (0.063-0.2 mm) for column
chromatography was from Merck.
2,6-Diethyl-2,3,6-trimethylpiperidin-4-ol (2). A 50% aque-
ous solution of NaOH (40 mL) was added to a solution of 197.3
g (1.0 mol) 2,6-diethyl-2,3,6-trimethylpiperidin-4-one57 (1) in
2-propanol (160 mL). Sodium borohydride (22.7 g, 0.60 mol)
was added, and the stirred mixture was heated to reflux under
nitrogen for 5 h. Water (80 mL) was then added, and the
mixture was refluxed for 5 min, cooled to 50 °C, and extracted
with toluene (2 × 100 mL). The organic layer was washed with
water (3 × 100 mL) and an aqueous NaCl solution (10%, 3 ×
80 mL), dried over MgSO4, and then evaporated on a rotary
evaporator to afford 2 (193.7 g, 97.2%) as a slightly yellow oil.
Capillary GC: six closely spaced peaks at 3.635 min (4.36%),
3.656 min (10.93%), 3.708 min (39.95%), 3.767 min (17.91%),
3.791 (19.40%), 3.842 min (6.73%), sum of all peaks ) 99.28%.
Anal. Calcd for C12H25NO (199.34): C, 72.31; H, 12.64; N, 7.03.
Found: C, 72.07; H, 12.62; N, 6.99. GC-MS(CI), m/e: six peaks
with identical spectra: 200 (42, [M + H]+), 182 (100). IR
(cm-1): 3341 s, 2964 s, 2933 s, 2879 s, 1460 s, 1373 s. 1H NMR
(300 MHz, CDCl3), mixture of diastereoisomers: δ ) 4.30-
3.60 (m, 1 H, -CH(OH)-), 1.80-0.80 (m, 24 H, 1 OH, 1 NH,

9976 Fischer et al.
Macromolecules, Vol. 38, No. 24, 2005
1 CH, 1 CH2, 3 CH3 , 2 C2H5). 13C NMR(75.38 MHz CDCl3): a
mixture of diastereoisomers, 48 poorly resolved signals clus-
tered between 69.79 and 27.29 ppm.
Scheme 3
4-Acetoxy-2,6-diethyl-2,3,6-trimethylpiperidin-N-ox-
yl (3 and 3a). Acetanhydride (55.9 g, 0.5475 mol) was added
dropwise over 30 min to a stirred solution of 4-(dimethyl-
amino)pyridine (0.56 g) in 2,6-diethyl-2,3,6-trimethyl-piperi-
din-4-ol (2) (99.67 g, 0.50 mol). The temperature during the
addition was kept below 45 °C. The mixture was then stirred
for another 30 min at 45 °C, cooled to room temperature, and
diluted with methyl-tert-butyl ether (250 mL). A solution of
sodium hydroxide (23.8 g, 0.595 mol) in water (125 mL) was
then added, and the organic layer was separated, washed with
water (30 mL) and an aqueous NaCl solution (10%, 2 × 30
mL), dried over MgSO4, and evaporated on a rotary evaporator
to afford 117.64 g (97.5%) of the O-acetylated 2 as a slightly
yellow oil. Capillary GC: six closely spaced peaks at 4.166 min
(6.80%), 4.196 min (9.89%), 4.233 min (32.73%), 4.249 min
(24.06%), 4.273 min (19.83%), 4.310 min (5.72%), sum of all
peaks ) 99.03%. Anal. Calcd for C14H27NO2 (241.38): C, 69.66;
H, 11.28; N, 5.80. Found: C, 69.55; H, 11.24; N, 5.75. GC-
MS(CI), m/e: six peaks with identical spectra (C14H27NO2
-CH(OAc)-, second diastereoisomer), 4.77 (q, J ) 6.6 Hz, 1H,
CH3CHC6H5), 2.35-0.35 (m, 25 H, 1 CH, 1 CH2, 4 CH3, 2
C2H5), 2.07 (bs, 3H, OCOCH3). 13C NMR (75 MHz, CDCl3), two
diastereoisomers, δ ) 170.96 (OCOCH3), 147.00 (CH3CHC6H5),
128.40 (C6H5), 127.19 (C6H5), 126.48 (C6H5), additionally ∼33
poorly resolved signals clustered between 84 and 8.47 ppm.
Acetic Acid 2,6-Diethyl-1-methoxycarbonyloxy-2,3,6-
trimethylpiperidin-4-yl Ester (36a). CuCl (1.98 g. 0.020
mol) was added to a solution of 3a (2.564 g, 0.010 mol) and
2-chloropropionic acid methyl ester (1.348 g, 0.0110 mol) in
ethyl acetate (20 mL) under nitrogen. Thereafter, N,N,N′,N′′,N′′-
pentamethyldiethylenetriamine (3.466 g, 0.020 mol) was added
dropwise over 20 min. The mixture was stirred at room
temperature for 4 h, diluted with ethyl acetate (50 mL), and
then filtered over a filtration aid (Hyflo). The filtrate was
washed with water (3 × 10 mL) and then with an aqueous
solution of EDTA disodium salt (1%, 3 × 10 mL) and water (3
× 10 mL). Drying over MgSO4 and evaporation on a rotary
evaporator afforded 3.40 g (99%) of 36a as colorless oil. Anal.
Calcd for C18H33NO5 (343.47): C, 62.95; H, 9.68; N, 4.08.
Found: C, 63.25; H, 9.75; N, 4.04. GC-MS (APCI), m/e: 344,
1
(241.38)) 242 (20, [M + H]+), 182 (100). H NMR (300 MHz,
CDCl3), mixture of diastereoisomers, δ ) 5.30-4.60 (m, 1H,
-CH(OAc)-), 1.95 (bs, 3H, OCOCH3), 1.80-0.40 (m, 23 H, 1
NH, 1 CH, 1 CH2, 3 CH3, 2 CH2CH3). 13C NMR(75.38 MHz
CDCl3), mixture of diastereoisomers, δ (OCOCH3) ) 170.64,
170.62, 170.43, 170.38, additionally 53 signals clustered
between 73.33 and 30.08 ppm.
1
(100, [M + H]+). IR (cm-1): 2987 m, 2951 m, 1735s. H NMR
Water (75 mL) and solid NaHCO3 (63.0 g, 0.75 mL) were
added to a solution of 60.34 g (0.250 mol) of the O-acetylated
2 in hexane (150 mL). A 40% solution of peracetic acid in acetic
acid (76.05 g, 0.40 mol) was then added dropwise to the stirred
mixture at 25-30 °C over 90 min. The red mixture was stirred
for 150 min at room temperature, then the organic layer was
separated, washed with water (100 mL), with aqueous Na2-
CO3 (1 M, 50 mL), and finally with a NaCl solution (10%, 2 ×
50 mL), dried over MgSO4, and evaporated on a rotary
evaporator to afford 63.45 g (99%) of 3 as a red oil which partly
solidified on standing. Capillary GC: eight closely spaced
peaks at 4.903 min (6.02%), 4.944 min (9.82%), 4.996 min
(5.21%), 5.050 min (17.71%), 5.090 min (35.47%), 5.108 min
(17.24%), 5.133 min (5.81%), 5.194 min (1.36%), sum of all
peaks ) 98.64%. Anal. Calcd for C14H26NO3 (256.37): C, 65.59;
H, 10.22; N, 5.46. Found: C, 65.60; H, 10.15; N, 5.40. GC-
MS (CI), m/e: eight peaks with identical spectra 256 (3, [M]+),
168 (24), 137 (67). IR (cm-1): 2975 m, 2938 m, 2882 w, 1738
s.
(300 MHz, CDCl3), two diastereoisomers, δ ) 5.20-5.09 (bs,
0.3H, -CH(OAc)-, first diastereoisomer), 4.98-4.82 (bs, 0.7H,
-CH(OAc)-, second diastereoisomer), 4.45-4.25 (m, 1H,
CH3CHCOOCH3), 3.72 (s, 3H, OCH3), 2.20-0.70 (m, 25 H, 1
CH, 1 CH2, 4 CH3, 2 C2H5), 2.06 (bs, 3 H, OCOCH3). 13C NMR-
(75.38 MHz CDCl3), two diastereoisomers, δ ) 174.72 (CH3-
CHCOOCH3), 170.72 (OCOCH3), 82.51, 80.48, 72.72, 68.34,
66.56, 66.08, 65.30, 65.10, 64.15, 63.66, 61.18, 60.80, 51.65
(COOCH3), 38.18, 37.09, 36.95, 36.49, 34.42, 31.86, 31.44,
29.65, 29.10, 27.39, 26.43, 25.49, 21.55, 21.36, 19.14, 18.88,
18.76, 12.77, 10.36, 8.38, 7.89, 7.74.
Kinetic Experiments. The solvent tert-butylbenzene
(Fluka, >99%) was purified following a known procedure.60
Rate constants kd were obtained from the appearance of the
nitroxyl radical ESR signal recorded in the presence of various
scavengers such as galvinoxyl, oxygen, and TMIO-15ND12 used
to remove irreversibly the released alkyl radicals.7,24 The
evolution of the nitroxyl radical signal was recorded with either
a EMX Bruker or a MS100 Magnettech EPR spectrometer. The
nitroxyl radical growth obeyed the expected first-order kinetics
and ended at the concentration expected for full decomposition
of the alkoxyamine. Measurements at different temperatures
yielded the frequency factors A and the activation energies Ea
by Arrhenius analysis.
Crystallization of the Diastereoisomer 3a. The diaste-
reoisomeric mixture 3 (30 g) was dissolved in pentane (15 mL)
and left at -30 °C for 15 h. The crystals (7.55 g) were filtered
off, washed with cold pentane, and recrystallized once again
from pentane (80 mL) at -30 °C to afford 7.05 g of the pure
isomer 3a as large red crystals, mp 86-89 °C. Capillary
GC: 1 peak at 5.090 min (98.6%).
Results
Acetic Acid 2,6-Diethyl-2,3,6-trimethyl-1-(1-phenyl-
ethoxy)piperidin-4-yl Ester (36). Tetrabutylammonium
bromide (0.075 g, 23.3 mmol) and CuIBr (0.032 g, 0.223 mmol)
were added to a solution of 5 g (0.0195 mol) of 3a in 10.3 g
(0.0970 mol) of ethylbenzene, and the mixture was deoxygen-
ated by four vacuum (0.1 bar)/argon cycles. An aqueous
solution (70%) of tert-butyl hydroperoxide (5.3 g, 41.2 mmol)
was then added dropwise at 45-50 °C over 10 min. The
mixture was stirred at 45-50 °C for 5 h, then diluted with 20
mL ethylbenzene, washed successively with 20 mL each of
aqueous Na2SO3 (20%), saturated NaHCO3, and saturated
NaCl, and then dried over Na2SO4. Evaporation of the solvent
on a rotary evaporator and chromatography on silica gel
column with hexanes-EtOAc (20:1) afforded 6.4 g (90.8%) of
36 as colorless oil. Anal. Calcd for C22H35NO3 (361.53): C,
73.09; H, 9.76; N, 3.87. Found: C, 72.87; H, 9.64; N, 3.85. MS-
(APCI), m/e: 362 (45, [M + H]+). IR (cm-1): 2973 s, 1736 s,
1452 m, 1367 s, 1239 s. 1H NMR (300 MHz, CDCl3), two
diastereoisomers, δ ) 7.39-7.22 (m, 5H, C6H5), 5.35-5.10 (m,
0.4H, -CH(OAc)-, first diastereoisomer), 5.10-4.88 (m, 0.6H,
Preparation of Alkoxyamines 36 and 36a. Com-
pounds 36 and 36a were prepared from the easily
available57 2,6-diethyl-2,3,5-trimethylpiperidin-4-one, 1,
which was transformed into nitroxyl radicals 3 and 3a
as depicted in Scheme 3.
Ketone 1 contains three asymmetric carbon atoms;
hence four diastereoisomers may be expected. GC-MS
analysis revealed only two isomeric peaks in a 2.2:1
ratio. However, four distinct C-carbonyl resonances in
the 13C NMR spectrum of 1 indeed indicated the
presence of four diastereoisomers.57 Nevertheless, given
the very narrow boiling range of 1 and its appearance
as a single spot on TLC, the separation of the isomers
by distillation or chromatography was not attempted.
Therefore, the formulas in Scheme 3, with the exception
of 3a, represent all possible diastereoisomers. Sodium
borohydride reduction of 1 followed by acetylation and

Macromolecules, Vol. 38, No. 24, 2005
C-ON Bond Homolysis Rate 9977
Table 2. Reestimated Activation Energies (Ea) and Rate
Constants (kd) at 120 °C, Universal Electrical (Polar)
Hammett Constants σI,n, Steric Constants Es,n, and Global
Steric Strain EN,n′ for the Nitroxyl Fragments of
Alkoxyamines 4′-45a
Scheme 4
b
c
d
e
alkoxyamines
Ea
kd
σI,n
0.50 -3.62f/-2.87f,h,i 123.2
0.60 -4.71 125.5
0.50 -3.62f/-2.87f,h,i 124.9
Es,n
EN,n′
4′
139.4f,g 7.1 × 10-5
5
138.5j
9.4 × 10-5
137.4f,g 1.3 × 10-4
peracetic acid oxidation afforded nitroxyl radical 3 as a
diastereomeric mixture. Crystallization of the isomeric
mixture of 3 from n-pentane afforded a single diaste-
reomer 3a. Its structure was determined by X-ray
analysis.61 Nitroxyl radical 3a was then transformed
into alkoxyamine 36 using Cu-catalyzed t-butylhydro-
peroxide-promoted coupling62 with ethylbenzene. The
slightly adapted protocol31 of Matyjaszewski was used
for the coupling of 3a with methyl chloropropionate to
afford 36a (Scheme 4).
6′
7
137.3k 1.4 × 10-4
0.07 -2.70
0.60 -6.21
0.60 -4.71
135.7
122.6
122.5
135.3
134.9
132.6
121.1
133.6
131.8
128.8
130.4
130.1
127.6
127.6
132.6
126.8
131.2
130.7
131.3
126.1
8
135.6j
135.5j
135.0l
134.6f
2.3 × 10-4
2.3 × 10-4
9
10
11
12
13
14
15
16
17
18′
19
20
21
22
23
24
25′
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
2.7 × 10-4 -0.01 -2.80
3.1 × 10-4 -0.02 -2.70
134.6k 3.1 × 10-4
0.08 -2.90
0.60 -6.21
0.00 -2.90
0.07 -2.99
0.18 -3.34
0.11 -3.32
0.09 -3.44
0.18 -2.74
0.15 -3.32
134.1j
3.6 × 10-4
133.6k 4.2 × 10-4
133.5f,m 4.3 × 10-4
133.3j
4.6 × 10-4
6.2 × 10-4
6.6 × 10-4
6.6 × 10-4
133.1k 4.9 × 10-4
132.3f
132.1j
132.1
131.6j
131.3j
130.9l
130.4l
129.7f
129.4
Kinetic Measurements. Because of the persistent
radical effect (PRE),5 measurements of kd were per-
formed in the presence of an excess of scavengers such
as galvinoxyl, TMIO-15ND12, or oxygen.7,24 Plots of
nitroxyl concentration evolution with time and Arrhe-
nius plots were already well exemplified in previous
works and do not need further reports and comments.7,24
As expected, frequency factors A of molecules 18, 25,
26, 33, and 34 lie in the range 1013-1015 s-1 and are
7.7 × 10-4 -0.04 -3.09
8.5 × 10-4
0.18 -2.74
9.6 × 10-4 -0.01 -3.09
1.1 × 10-3 -0.01 -3.41
1.4 × 10-3 -0.06 -2.95
1.5 × 10-3
0.06 -3.95f
128.7k 1.9 × 10-4
0.07 -3.70f/-4.45f,h,n 127.0
128.6k 1.9 × 10-3 -0.02 -3.59
129.1
123.9
125.4
123.5
centered ca. 1014 s-1 23-25,27,32,36,38,39
Therefore, the 33
.
128.4j
2.1 × 10-3
0.18 -3.31
0.11 -5.30
0.18 -3.31
values of A measured by Marque et al.23-25,36,38 give an
averaged value of 2.4 × 1014 s-1, and this value is used
to estimate and to reestimate the values of Ea (Tables
1 and 2). While this work was in progress, the kd of 11
and 20 were published by Fukuda et al.63 and Studer
et al.,54 respectively, and the values obtained by the
authors agree well with these reported values. The Ea
of 36a (135 kJ/mol, Table 1) falls close to the expected
value (132 kJ/mol) when applying the increment
(∆Ea,PhEtfEEst ) +5.7 kJ/mol) given by Marque et al.24
Taft)Ingold Approach. Recently, one of us46 showed
for 24 nitroxyl moieties that the effects of the substit-
uents on the nitroxyl moiety are rationalized in terms
of the Taft-Ingold equation (eq 1) where σI,n and Es,n
are the universal electrical (polar inductive/field effect)
and steric substituent constants for a given alkoxyamine
n. Thus, the ring substituent and ring size effects in
molecules 4′-45 were analyzed with eq 1. All the values
of Ea given in Table 2 were reestimated with an
averaged frequency factor A of 2.4 × 1014 s-1 and for
the PhEt group (R group in Scheme 2) as released alkyl
radical. For molecules 18′ and 25′, the values of Ea for
a nitroxyl moiety carrying a PhEt group were estimated
using the increment64 (∆Ea,PCNfPhEt) of +18.0 kJ/mol to
give 132.3 and 129.7 kJ/mol, respectively. The values
of kd at 120 °C, σI,n, Es,n, and EN,n′ (global steric strain
of nitroxyl moiety given by eq 2) are gathered in Table
2.
128.1k 2.3 × 10-3
128.0j
127.5
127.2
126.7
2.3 × 10-3
2.7 × 10-3
0.00 -3.95f/-3.19f,h,o 127.5
3.0 × 10-3 -0.02 -3.95f/-3.19f,h,p 127.7
3.5 × 10-3 -0.01 -3.70
127.0
127.5
126.7k 3.5 × 10-3 -0.03 -3.59
126.4
125.7l
3.8 × 10-3
0.00 -3.95f/-3.19f,h,q 126.4
4.7 × 10-3 -0.01 -3.49
126.0
126.2
124.3
119.7
124.8
125.7k 4.7 × 10-3 -0.02 -6.20
124.3k 7.2 × 10-3 -0.02 -4.20
124.2j
7.4 × 10-3
0.18 -4.85
123.8j
8.4 × 10-3 -0.04 -4.38
123.7k 8.7 × 10-3
0.07 -4.70f/-5.45f,h,r 122.0
123.1
122.8j
122.4
122.2j
1.1 × 10-2 -0.01 -4.70
1.3 × 10-2
0.15 -5.06f/-3.81f,h,s 118.7
1.4 × 10-2 -0.04 -4.38
123.2
a Ea and EN,n′ in kJ/mol and kd in s-1 at 393 K. b Ea was
reestimated with a frequency factor of 2.4 × 1014 s-1 (see text)
and for an alkoxyamine carrying a PhEt as leaving alkyl radical,
unless otherwise mentioned. c Estimated with eqs 3-5. d Esti-
mated with eqs 6 and 7. e Estimated with eq 2. f See text. g Given
i
in ref 75. h Italized values are for triangles in Figure 1. Es,4′ and
Es,6′ were estimated with r(C5) and r(C6′′). j Given in ref 45.
k Given in ref 40. l Given in refs 42 and 43. m Given in refs 41 and
n
o
70. Es,27 was estimated with r(C6′). Es,32 was estimated with
r(C6). p Es,33 was estimated with r(C6). q Es,36 was estimated with
r(C6). r Es,42 was estimated with r(C6′). s Es,44 was estimated with
r(C7).
the nitroxyl function. All individual σI constants were
not known, and the missing ones were estimated67 with
eqs 4 and 5. The values of Es,n were given by eq 6
assuming a ) b ) 1 and ꢀ ) 0, because all molecules
were cyclic and therefore exhibited symmetry or
EN,n′ ) Ea - 25.0σI,n
(2)
i)6
pseudosymmetry.46 The values of Es and Es were
given53 by eq 7 and the individual steric constant r(Ri)
with i being the rank related to the value of r(Ri) i.e.,
its size (the more negative r(Ri) is, the larger Ri is) and
r1 g r2 g r3. All individual σI(Ri) and r(Ri) values are
listed in Table 3. Unknown values of r(Ri) were esti-
mated either by analogy68 like for r(CMe2CN)69 and
r(CH2OCH2Ph) or assuming that the value of kd obeyed
eq 1 like for r(C7), r(C7′), r(C8), r(C8′), r(COOMe), and
r(CH2OSiMe2t-Bu). The value of r(C7′′′) was estimated
A
B
σI,n
)
σ (R )
I i
(3)
∑
i)1
σI,R CH ) 0.416σI,R - 0.0103
(4)
(5)
1
2
1
σI,R R CH ) 0.297ΣσI,R + 0.00482
1
2
The values of σI,n were given46 by eq 3 where σI(Ri) is
the individual Hammett constant65,66 σI of each R group
bonded to the tetrahedral carbon in the positions R to

9978 Fischer et al.
Macromolecules, Vol. 38, No. 24, 2005
Table 3. Individual Steric Constants r and Electrical Universal (Polar) Constants σL
groupa
rb
σI(Ri)c
groupa
rb
σI(Ri)c
CR2(CH2)2CR2
0.31d,e
0.20d,f,h
0.22d,f
0.22d,f
0.22d,e
0.19d,f
0.19d,f
0.13d,f
0.11d,e
0.11d,e
0.11d,f
0.11d,f
0.09d,f
0.09d,f
-0.02
0.54i
0.11k
0.03l
-0.02e
0.12m
0.04n
0.15o
0.37e
0.10p
0.04q
0.02s
0.02t
0.01u
CR2CH2C(O)NHCHMeCR2
0.02d,f
0.00d,f
0.00
0.19g
CR2NMeC(O)CR2
CR2(CH2)2CHOHCH2CR2
0.04j
CR2CH2C(O)CH2CR2
CR2CH2CHOHCH2CR2
CR2(CH2)3CR2
CR2(CH2)3C(O)CH2CR2
CR2(CH2)2C(O)(CH2)2CR2
CR2(CH2)2COCH2CR2
CR2CH2NRCOCR2
CR2CH2COCMeHCR2
CR2CH2CHOAcCHMeCR2
CR2CH2CHOHCHMeCR2
CR2(CH2)3CHOHCH2CR2
CR2(CH2)2CHOH(CH2)2CR2
Me
(CH2)4
-0.01
-0.02
0.11
-0.04d
-0.06f
-0.15d
-0.27d
-0.30f
-0.38
-0.50f
-0.50f
-1.40e
-1.55f
-2.10v
CH2OH
(CH2)5
-0.02
-0.02
0.00e
-0.01
0.11
(CH2)6
CH2OSiMe2t-Bu
Et
CH2OMe
CH2OCH2Ph
Ph
COOMe
CMe2CN
0.11f,r
0.12
0.64
0.14
a Groups attached to the nitroxyl moiety. b Individual steric constants r(Ri) are given in ref 53 unless otherwise mentioned. c Individual
polar constant σI(Ri) are given in refs 65 and 66 unless otherwise noted. d In text, r(C5) is used for CR2(CH2)2CR2; r(C5′) for (CH2)4; r(C5′′)
for CR2NMeC(O)CR2; r(C6) for CR2(CH2)3CR2, CR2CH2C(O)CH2CR2, and CR2CH2CHOHCH2CR2; r(C6′) for CR2CH2NRC(O)CR2,
CR2CH2C(O)CMeHCR2 CR2CH2CHOAcCMeHCR2, and CR2CH2CHOHCMeHCR2; r(C6′′) for (CH2)5; r(C7) for CR2(CH2)2CHOHCH2CR2;
r(C7′) for CR2(CH2)2C(O)CH2CR2 and CR2CH2NHC(O)CH2CR2; r(C7′′) for (CH2)6; r(C7′′′) for CR2CMeHNHC(O)CH2CR2; r(C8) for
CR2(CH2)2CHOH(CH2)2CR2 and CR2(CH2)3CHOHCH2CR2; and r(C8′) for CR2(CH2)2C(O)(CH2)2CR2 and CR2(CH2)3C(O)CH2CR2. e Given
g
in ref 46. f See text. σI,CH NHAc ) 0.09, assuming σI,CH C(O)NHMe ≈ σI,CH2C(O)NMe given by eq 4 and σI,C(O)NMe ) 0.28. h See text for the
2
2
2
PhCH2 group. i Assuming σI,C(O)NMeH ≈ σI,C(O)NMe ) 0.228 and σI,NMeAc ≈ σI,NHC(O)Et ) 0.26. j σI is estimated with σI,CH CH CHOHMe ) 0.01
2
2
2
k
l
given by eq 4 and σI,CH CHOHMe ) 0.03. σI,CH Ac ) 0.11 is estimated from eq 4 and σI,Ac ) 0.30. σU ) 2 × (σI,CH Ac/2); see ref 46. σI,CH CHOHMe
) 0.03, σI ) 2 × (σI,CH CHOHMe/2); see ref 46. m σI is estimated with σI,CH Ac ) 0.11 (see footnote k) and σI2,(CH ) Ac ) 0.01 given by eq 4.
2
2
2
2
2
3
/2); see ref 46. o σI2 is estimated with σ
n
σ
) 0.04 given by eq 4 with σ
) 0.11 (see footnote k), and σU ) 2 × (σ
I,(CH2)2Ac
I,CH2Ac
I,(CH2)2Ac
I,CH2Ac
p
) 0.11 (see footnote k) and σI,(CH ) Ac ) 0.04 (see footnote n). σI,CMeHAc ) 0.09 is estimated with eq 4 and σI,Ac ) 0.30. σI ) σI,CMeHAc/2 +
2
2
σI,CH Ac/2; see ref 46. q σI is estimated with σI,CH CHOAcMe ) 0.04 given by eq 4 and σI,CMeHCHOAcMe ) 0.04 given by eq 5. σI ) σI,CH CHOAcMe/2
2
2
2
+ σI,CMeHCHOAcMe/2. r Assuming σ
≈ σ
.
s σI,CMeHCHOHMe ) 0.01 is estimated with eq 5 and σI,CHOHMe ) 0.04, and σ
I,CH2OMe
I,CH2OCH2Ph
I,CH2CHOHMe
) 0.03. σI ) σI,CMeHCHOHMe/2 + σI,CH CHOHMe/2; see ref 46. t σI is estimated with σI,(CH ) CHOHMe ) -0.01 given by eq 4 and σI,CH CHOHMe
)
2
2
3
2
0.03. σI,CH CH CHOHMe ) 0.01 given by eq 4 and σI,CH CHOHMe ) 0.03. σI ) 2 × (σI,CH CH CHOHMe/2); see ref 46. v Estimated by analogy with
u
2
2
2
2
2
the value of CHF2 given in ref 69.
kJ/mol.71 For alkoxyamines 4′ and 6′ (pentagons in
Figure 1), assuming a small difference of polarity
between the methyl and the benzyl group,72 values of
r(C5′′) and r(C6′′) were used to determine the values of
Es,4′ and Es,6′. The value of r(C5′′)73 was preferred to
r(C5) because it yielded a better fit (vide infra).74 The
Ea values of 4′ and 6′ were measured by Aldabbagh et
al.75 for the polystyryl (PS) derivatives. For TEMPO-
PS and SG1-PS derivatives, it was shown by Fukuda
et al.20 and Bertin et al.,16 respectively, that the PS
chain increases the value of kd by a factor of 2 at best.
Therefore, the values of kd for the PS derivatives of 4′
and 6′ were divided by 2, and Eas reestimated with the
averaged frequency factor A (vide supra). For alkoxy-
amines carrying polar groups in positions 2 and 6 on
the ring of the nitroxyl fragment (crossed hexagons in
Figure 1), the values of Es,n were estimated with r(C6)
and with r(COOMe) for 5, 8, 9 and 13, r(CH2OMe) for
16 and 40, and r(CH2OSiMe2t-Bu) for 21, 41 and 45.
The lower values of EN′ for 5, and 9 than for 11 (Table
2) indicate that r(COOMe) is larger than r(Me), and
assuming that the nitroxyl fragment adopts a preferred
conformation, a value of -1.55 was estimated for
r(COOMe). For alkoxyamines 16 and 40, it was assumed
that the polar effects of the CH2OMe and CH2OCH2Ph
Figure 1. Log kd vs eq 1 for alkoxyamines 4′-45 based on
various cyclic nitroxyl fragments, (>) for five-membered rings,
(`) for six-membered rings and Es estimated with r(C6), (")
for six-membered rings and Es estimated with r(C6′), (hexagon
with a cross) for six-membered rings carrying polar groups in
positions 2 and 6, (f) for six-membered rings carrying groups
capable of IHB, ([) for seven-membered rings, (b) for eight-
membered rings, (2) for log kd estimated with italicized values
of Es,n
.
by removing 0.11 from r(C7′) by analogy to the differ-
ence between r(C6) and r(C6′).
groups were very close, i.e. σI(CH2OCH2Ph)
≈
Es,n ) aEAs + bEBs + ꢀ
(6)
σI(CH2OMe),65-67 as well their steric demands, i.e.,
r(CH2OCH2Ph) ≈ r(CH2OMe), by virtue of the minimum
steric interaction (MSI) principle.68,76 Since the O-Si
bond is longer than the O-C bond,76 the steric demand
of the OSiMe2t-Bu group was expected to be smaller
than that of the methoxy group; that is, r(CH2OSiMe2-
t-Bu) ) -0.30 was the best fitting value of eq 1. For
alkoxyamines carrying groups capable of intramolecular
hydrogen bonding (IHB) in positions 2 and 6 on the ring
of the nitroxyl fragment (stars in Figure 1), the Es,n
values were estimated with r(C6) and r(CH2OH) for 19,
Es(CR1R2R3) ) -2.104 + 3.429r1(R1) +
1.978r2(R2) + 0.649r3(R3) (7)
For alkoxyamines 15 and 18′ (pentagons in Figure 1),
the values of r(C5) and r(Ph) estimated by Marque were
used to determine the values of Es,15 and Es,18′. The
value of Ea for 15 was measured by Cameron et al.,41,70
but it was reestimated with the averaged frequency
factor (A ) 2.4 × 1014 s-1) and was incremented by 2.7

Macromolecules, Vol. 38, No. 24, 2005
C-ON Bond Homolysis Rate 9979
22, 29, and 31.77 For alkoxyamines 7, 10, 11, 23 - 25′,
27, 34, 37, 42, and 43 based on a six-membered ring
nitroxyl fragment without substituents in position 3
(filled hexagons in Figure 1), the values of Es,n were
estimated with r(C6). For alkoxyamines 10, 23, 24, and
37 (filled hexagons in Figure 1), r(C4), r(C5′) r(C6′′) were
used for the 5-, 6- and seven-membered rings as cyclic
substituents in the positions R, respectively. For
alkoxyamines 26, 32, 33 and 36 based on a six-
membered ring nitroxyl fragment carrying one substitu-
ent in position 3 (sp3 hybridized carbon for the methyl
group,78 open hexagons in Figure 1), the Es,n values were
estimated with r(C6′). Because of the assumption made
in a previous work (see discussion),46 Es,n values for
alkoxyamines 32, 33, and 36 were reestimated using
r(C6), and Es,n values for alkoxyamines 7, 27, and 42
were reestimated using r(C6′). Both families of mol-
ecules are the triangles and italicized numbers in Figure
1, and they have Es,n italicized values in Table 2. For
alkoxyamines based on a seven-membered ring nitroxyl
fragment (diamonds in Figure 1), it was assumed that
17 and 39 obeyed eq 1, and thus values of 0.13 and 0.00
were estimated for r(C7) and r(C7′) for a seven-
membered ring containing one sp2 hybridized carbon
atom (17) and a seven-membered ring with all sp3
hybridized carbon atoms (39), respectively. These values
of r(C7) and r(C7′) were applied to molecules 20 and 30
and molecule 38, respectively. It has to be mentioned
that 30 and 38 are outliers (see leveled steric effect
section), and for the sake of simplicity, 38 is not reported
in Figure 1 because it deviates markedly from the
regression line (coordinates are 5.52 for the x axis and
-2.33 for the y axis; see Figure 1SI in Supporting
Information). For molecule 44, r(C7′′′) was used to
estimate Es,44 (vide supra). For alkoxyamines based on
a eight-membered ring nitroxyl fragment (circles in
Figure 1), because the kd values are known only for 12,
14, 28, and 35, it was assumed that 14 and 28 obeyed
eq 1, and thus values of 0.19 and 0.09 were estimated
for r(C8) and r(C8′) for a eight-membered ring contain-
ing a sp2 hybridized carbon atom (12 and 14) and a
eight-membered ring with all sp3 hybridized carbon
atoms (28 and 35), respectively.
Discussion
General Comments on the Arrhenius Param-
eters and the Taft)Ingold Equation. From the
literature,79 A values around 1013 to 5 × 1013 s-1 are
expected for the bond homolysis of small molecules i.e.,
no activation entropy effect, whereas larger A values
(1017-1018 s-1) are expected for the fission of molecules
in two large groups. These high A values are due to an
increase of activation entropy because the number of
freedom for rotation increases.79 However, the homolysis
of alkoxyamine into two large groups i.e., the alkyl and
nitroxyl radicals, requires the passage from sp3-hybrid-
ized nitrogen atom in the initial state to sp2-hybrized
nitrogen atom in the transition state (from the Ham-
mond postulate, for endothermic reaction the structure
of the transition state resembles to the structures of the
products i.e., nitroxyl and alkyl radicals). Such rehy-
bridization is energetically expensive in hindered mol-
ecules such as R,R′-tetrasubstituted alkoxyamines, and
this should decrease the activation entropy. Conse-
quently, the A values are expected in the range 1013-
1015 s-1 as observed (Table 1). Furthermore, rings can
adopt many conformations, and the entropic cost in-
creases with the number of conformation “to freeze” in
the transition state, and the A values decrease. Up to
now, no logical variations of A values have been
observed, thus for the sake of simplicity, it is assumed
that the averaged frequency factor (A ) 2.4 × 1014 s-1
)
holds for all alkoxyamines. The bond strength (enthalpic
term), the steric hindrance and the polarity of the bond
are all included in the activation energy Ea. In general,
eq 1 is valid only for a family of compounds.80 For
practical reasons (vide supra), one seeks a general
relationship for all cyclic alkoxyamines, and the devel-
opment of constants for the cyclic strain makes it
possible to take into account the entropic effect due to
the number of possible ring conformations. Hence, a
relationship taking into account all rings whatever the
size is possible, and this makes it possible to detect the
entropic changes due to the substituents (vide infra).
Five-Membered Ring Nitroxyl Fragments. The
steric effects of various groupssMe,46 Et,46 Ph, cyclo-
hexyl, and CMe2CNsare well accounted for by eqs 3-8
(pentagons in Figure 1). For five-membered ring nitroxyl
fragment based alkoxyamines 4′ and 6′ close to the line,
r(C5′′) has to be used instead of r(C5),73 which means
the cyclic strain or the number of conformations (vide
infra) is modified by the substitution or hybridization
sp2 or sp3 in positions 3 and/or 4, as observed for six-
membered ring nitroxyl fragment based alkoxyamines
(see below). It is probable that the buttressing effect
mentioned by Aldabbagh et al. does not occur.75
Except for 6′, 5, 8, 13 19, 22, 29-31, 37 and 38, the 31
alkoxyamines left lie very close (eq 8) to the regression
line given by eq 1. In eq 9, the regression is performed
on all the known alkoxyamines based on cyclic nitroxyl
fragments (31 alkoxyamines presented in this work and
12 alkoxyamines previously presented by Marque)46 and
exhibit very good statistical parameters. Very good
statistical parameters were again obtained when the
biparameter regression (eq 10) was performed including
alkoxyamines based on linear nitroxyl fragments.
Six-Membered Ring Nitroxyl Fragments. In a
previous work,46 it was mentioned that the sp2 or sp3
hybridization of the C atom in position 3 modifies the
values of the cyclic strain constant r(C6). Assuming that
r(C6′) should be applied for a ring containing a sp2
hybridized C atom,46 alkoxyamines 7, 27, and 42
(triangles in Figure 1) should lie close to the regression
line, like 26, whereas they lie close to the regression
line when r(C6) is applied (filled hexagons in Figure 1).
Reciprocally, assuming that r(C6) should be applied for
a ring with all sp3 hybridized atoms,46 alkoxyamines 32,
33, and 36 (triangles in Figure 1) should lie close to the
regression line, like for example 11, 34, and 43, whereas
they lie close to the regression line when r(C6′) is
applied (open hexagons in Figure 1). Consequently, for
log(kd/s-1) ) -2.85((0.16)σI,n - 0.83((0.04)Es,n
-
5.72((0.11) (8)
R2 ) 0.96; s ) 0.13; N ) 31; F30,0.01% ) 340
log(kd/s-1) ) -2.81((0.15)σI,n - 0.83((0.03)Es,n
5.73((0.11) (9)
R2 ) 0.95; s ) 0.15; N ) 43; F42,0.01% ) 383
log(kd/s-1) ) -2.91((0.13)σI,n - 0.85((0.03)Es,n
-
-
5.81((0.09) (10)
R2 ) 0.96; s ) 0.14; N ) 53; F52,0.01% ) 591

9980 Fischer et al.
Macromolecules, Vol. 38, No. 24, 2005
Therefore, such activation entropy effect should be
observed through changes in the values of the frequency
factor A. Assuming that the small polar effect due to
the hydroxyl group in position 4 on the ring of the
nitroxyl fragment of 46 is contained in the activation
energy Ea,11sthat is, in the polar term88 of the BDE-
(C-O)sand assuming that the substituent in position
4 has no steric effect (see Table 3), the frequency factor
A (not estimated) for the C-ON bond homolysis of 11
should be close to the frequency factor observed for the
homolysis of TEMPO-PhEt; that is, A11 ) 2.5 × 1014
s-1. Assuming the methyl group in position 3 on the ring
of the nitroxyl fragment in 46 has an influence only on
the activation entropy in TS, that is, Ea,46 ) Ea,11
)
134.6 kJ/mol and with kd,46 ) 1.6 × 10-3 s-1 (see
above),81 one should expect a value of 1.2 × 1015 s-1 for
A46. This is a reasonable value for bond homolysis and
is 5 times higher than that of 11. However, despite a
given error of a factor two for the values of A, it would
have been difficult to observe this 5-fold increase
because of the error compensation effect of the activation
parameters,89 that is the same rate constants can be
estimated from different (Ea, A) couples. It has to be
mentioned that several authors have observed frequency
factors A above 1015 s-1 but with an increase of Ea that
undermines any unambiguous interpretation of the
results.90 The steric influence of the monocyclic substit-
uents in molecules 10, 23, and 24 is well accounted for
by the cyclic steric constants (Table 3) of Fujita.53 The
upward deviation observed for 37 might be due to the
presence of two cyclohexyl groups generating some
deformations of the nitroxyl ring and making the
cyclohexyl groups bulkier than expected (r(C6′′)) but
smaller than the ethyl group. Molecules 5, 8, 9, 13, 16,
21, 40, 41, and 45 (crossed hexagons in Figure 1) were
plotted separately because of the polar effect due to the
presence of the COOMe, CH2OMe, CH2OCH2Ph, and
CH2OSiMe2t-Bu groups. The closeness of molecules 16,
21, 40, 41, and 45 to the regression line shows that the
steric and polar effects of the CH2OMe, CH2OCH2Ph,
and CH2OSiMe2t-Bu groups are well accounted for by
eqs 2-8. Furthermore, it also shows that the configu-
ration cis-trans around the nitroxyl moiety has no
significant influence on kd. For molecules 5, 8, 9, and
13, the analysis is made more difficult because the ester
group can adopt many conformations, which generate
different steric strains (Janus group), and hence
r(COOMe) can have different values. Assuming the
bulky ester groups prefer pseudoequatorial positions to
pseudoaxial positions, molecule 9 with the cis relation-
ship for the ester groups should provide the “normal”
value for r(COOMe); that is, r(COOMe) ) -1.55.
However, using this value to estimate Es,n for 5, 8 and
13 makes these molecules outliers. The downward
deviation of 5 is probably due to a larger number of
conformations to be frozen in TS (see above) because of
the trans relationship of the ester group in 5 always
involving an ester group in pseudoaxial position, whereas
it cannot occur for 9. The downward deviations of 8 and
13 are discussed in the leveled steric effect section.
Molecules 19, 22, 29, and 31 (stars in Figure 1) are
discussed in the intramolecular hydrogen bonding sec-
tion.
Figure 2. X-ray structure of 3a.
Scheme 5
a six-membered ring, the hybridization of the ring atoms
has no effect on kd but the effect of the substituent
position is striking; that is, a substituent in position 4
has no significant steric effect whereas a substituent
in position 3 increases kd. This is highlighted by the
marked increase of kd for 46 (Scheme 2); e.g., kd,46 (kd
) 1.6 × 10-3 s-1 81
)
is roughly 5 times higher than kd,11
(kd ) 3.1 × 10-4 s-1 in Table 2). It is well-known that
stereomutations in alkoxyamines based on cyclic ni-
troxyl fragments are many and are very complicated
processes.82-87 Thus, it can be assumed that the ring
adopts several conformations which have to be “frozen”
in TS, which is costly in terms of activation entropy;
i.e., ∆Sq is small and thus kd decreases. On the grounds
of the endothermic C-ON bond homolysis and the
Hammond postulate, one can assume a late TS and a
structure and a conformation for the nitroxyl fragment
close to that of the nitroxyl radical. Assuming that the
nitroxyl fragment in the alkoxyamine adopts the same
conformations as those of the nitroxyl radicals, the X-ray
structure of the nitroxyl radical 3a (Figure 2 and
Scheme 5), i.e., the most stable conformation in the
crystal state and the one assumed in solution, reveals
the two ethyl groups in pseudoequatorial position, as
expected for bulky groups, the methyl group in position
3 in equatorial position (trans relationship with the
ethyl groups), and the acetoxy group in position 4 in
axial position (trans relationship with the ethyl groups).
A chair a chair chemical exchange (form A a form B,
Scheme 5) in 3a leads to the disfavored form B because
of four strong 1,3-syn interactions (Et T Et, 2Et T H,
and Me T H interactions). In the absence of substituent
in position 3, a chair a boat chemical exchange is
possible whereas such chemical exchange (form C,
Scheme 5) is forbidden by the presence of the methyl
group in position 3, that is, form C exhibits strong Me
T Me 1,2- and 1,3-syn interactions. Therefore, if con-
formers such as forms B and C do not exist in TS
because of too high steric strain, then fewer conforma-
tions need to be “frozen” in TS, and thus ∆Sq is less
small.
Seven-Membered Ring Nitroxyl Fragments. Mol-
ecules 30 and 38 are discussed in the leveled steric effect
section. The difference of 6 kJ/mol between the global
steric strain of E′N,17 and that of E′N,39 suggests that

Macromolecules, Vol. 38, No. 24, 2005
C-ON Bond Homolysis Rate 9981
the hybridization sp2 or sp3 of the ring carbon strongly
modifies the homolysis of the C-ON bond. That is, the
presence of a sp2 carbon on the ring might either induce
more strain in the ring or forbid the conformations
required in TS for the cleavage to occur. It is worthwhile
to mention that E′N,39 is the same as E′N for the
alkoxyamine based on the di-tert-butyl nitroxyl frag-
ment,46 which means no more cyclic strain for a seven-
membered ring. For alkoxyamine 44, when Es,44 is
estimated with r(C7) the molecule deviates strongly
(triangle in Figure 1 and italicized value in Table 2),
whereas the molecule lies on the regression line (Figure
1) when r(C7′′′) is applied; i.e., r(C7′′′) is smaller than
r(C7) by 0.11. This difference of 0.11 is the same
difference as that observed between r(C6) and r(C6′),
which would mean the same entropic effect observed
with 26 is also observed with 44. Hence, the presence
of a methyl group probably limits the number of
conformers to “freeze” in TS, and thus, ∆Sq is less small
and kd higher (vide supra). Note that the effects of the
mono- or multisubstitution and sp2 hybridization on
positions 3, 5, and 6 are not yet fully understood.
Eight-Membered Ring Nitroxyl Fragments. The
closeness of alkoxyamines 12, 14, 28, and 35 to the
regression line (circles in Figure 1) shows that the
values of r(C8) and r(C8′) depend only slightly on
positions 4 (E′N,12 and E′N,28) or 5 (E′N,14 and E′N,35) of
the subtituents on the ring but depend more on the sp2
(E′N,12 and E′N,14) or sp3 (E′N,28 and E′N,35) hybridization
of the carbons on these positions. The same explanation
as the one given for alkoxyamines based on seven-
membered ring nitroxyl fragments (vide supra) should
account for the difference of 3.5 and 2.9 kJ/mol between
E′N,12 and E′N,28 and between E′N,14 and E′N,35, respec-
tively. Unlike what was expected, i.e., no cyclic strain
for rings with more than seven atoms (vide supra),
alkoxyamines 12, 14, 28, and 35 exhibit larger E′N,n
than their seven-membered ring homologues 17 and 39;
that is, Ea,12 or 14 and Ea,28 or 35 are smaller than Ea,17
and Ea,39, respectively. Such difference might be due
either to increasing cyclic strain effects with the increase
of the ring size or to more conformations to be “frozen”
in TS because of a larger ring size, that is, the confor-
mational effects overbalance the relief of the cyclic
strain.
Leveled Steric Effect. The downward deviations
observed ca. 9 kJ/mol for 8 and 13, ca. 8 kJ/mol for 30,
and ca. 14.7 kJ/mol for 38 are too large to be ascribed
to the experimental error (ca. 2 kJ/mol). Moreover, the
estimated Es,8, Es,13, and Es,38 values are below -6, the
value from which the leveled steric effect26,91 may occur.
Because ester groups can adopt many conformations,
the replacement of the methyl groups of 9 by more
sterically demanding ethyl groups in 13 may push the
ester groups to adopt other conformations which induce
relief strain, and for example, choosing92 r(COOMe) at
-0.42 shifts 8 and 13 toward the regression line. Both
the deviating alkoxyamine 38 and the Es,38 smaller than
-6 (Es,38 ) -6.2; see Table 2) point out the presence of
the leveled steric effect. To move 38 closer to the
regression line, the leveled steric effectsdue to the
replacement of four methyl groups in 39 by four more
sterically demanding ethyl groupssmight be balanced
by either an increase of the cyclic steric strain (r(C8′) )
0.31, Es ) -4.1), or conformation changes which make
the ethyl groups of 38 no more sterically demanding
than the methyl groups of 39 (Es ) -4.2). Although Es,30
is above the threshold value but close to it, the leveled
steric effect accounts for the observed reactivity. As-
suming either a value of 0.31 for r(C8) or only two active
ethyl groups either on each side (Es ) -3.80) or on the
same side (Es ) -4.2) the leveled steric effect may be
balanced.
Intramolecular Hydrogen Bonding (IHB). A few
years ago,36 one of us showed the importance of IHB
for the kd of alkoxyamines containing nitroxyl fragments
carrying groups capable of IHB. With polar and steric
(r(CH2OH) ) -0.06)77 effects accounted for by eq 1, the
upward deviations (ca. 7 kJ/mol) observed for 19, 22,
29, and 31 are too large to be ascribed to experimental
errors (ca. 2 kJ/mol), and they are due to the potential
IHB between the hydroxyl groups in positions 2 and 6
and the nitroxyl moiety. This IHB provokes a stabiliza-
tion of the nitroxyl radical and thus of TS (Hammond’s
postulate) and hence increases kd. It can be argued that
r(CH2OH) is smaller than the estimated value, but
values smaller than -0.5 given for r(CH2OMe) are
unrealistic. Therefore, assuming the value of -0.5 for
r(CH2OH), one would observe (not shown) upward
deviations for 19 and 22 and downward deviations for
29 and 31, which means the presence of IHB for 19 and
22 and some others effects that slow the homolysis for
29 and 31. In contrast to previous observations,36,46
alkoxyamines 19, 22, 29, and 31 yielding nitroxyl
radicals capable of IHB do not always exhibit higher kd
than alkoxyamines 16, 21, 40, 41, and 45 yielding
nitroxyl radical carrying protected hydroxyl groups
(Table 3); that is, kd(19) ≈ kd(16) and kd(21),93 kd(22) >
kd(16) and kd(21), kd(29) < kd(41), and kd(31) < kd(40)
and kd(45). Such differences are merely accounted for
by the steric effect caused by the difference in bulkiness
between the CH2OH (r ) -0.06) and CH2OMe (r )
-0.50), CH2OCH2Ph (r ) -0.50), and CH2OSiMe2t-Bu
(r ) -0.30) groups (Table 3).
In a previous work, it was observed that nitroxyl
radicals capable of IHB generally exhibit higher EPR
nitrogen hyperfine coupling constants aN than the
corresponding nitroxyl radicals carrying protected hy-
droxyl groups, that is, nitroxyl radicals 19•, 22•, 29•, and
31• exhibit higher aN than nitroxyl radicals 16•, 40•, 21•,
41•, and 45• (Scheme 6) i.e., aN,31• ) 1.523 mT > aN,40•
) 1.430 mT and aN,45• ) 1.437 mT, aN,29• ) 1.530 mT >
aN,41• ) 1.380 mT, and aN,22• ) 1.520 mT > aN,16• ) 1.473
mT and aN,21• ) 1.516 mT.45,94 Hence, the higher aN
observed is in rather good agreement with a IHB
stabilization of the TS of 19, 22, 29, and 31.
Concluding Remarks
This study confirms many observations that have
been done by us and others, such as the following.
•kd increases with the increasing ring size of the
nitroxyl fragment up to seven members; i.e., the cyclic
strain decreases along the series below:
The ring effects of the nitroxyl moiety can be sum-
marized as follows: (i) five-membered rings exhibit

9982 Fischer et al.
Macromolecules, Vol. 38, No. 24, 2005
Scheme 6
strong cyclic strain that pulls back the substituents in
R-positions and thus relief of the steric hindrance
around the nitroxyl function is observed; (ii) the six-
membered ring exhibits less cyclic strain than the five-
membered one, and hence, the steric hindrance around
the nitroxyl function is increased but sensitive to
substitution on the 3-position of the ring; (iii) the seven-
membered ring is large enough for its weak cyclic strain
to match the steric hindrance of two tert-butyl groups
bonded to the nitrogen atom, and the importance of the
cyclic strain relief is sensitive to the hybridation of
carbons 4 or 5 of the ring, and to the presence of
substituents on positions 3 and 6. The steric strain
would be smaller for rings composed of more than seven-
members, as exemplified with the alkoxyamines based
on eight-membered ring nitroxyl fragments.
•IHB increases markedly the values of kd by stabiliz-
ing TS through stabilization of the nitroxyl radical.
•Electron-withdrawing groups attached to the ring of
the nitroxyl fragment decrease markedly the values of
kd: the closer they are to the nitroxyl moiety, the
smaller kd is.
•The leveled steric effect also occurs for values of Es,n
below -6 in alkoxyamines carrying cyclic nitroxyl
fragments like those for alkoxyamines carrying linear
nitroxyl fragments.
•Substitution on the carbon atom of the position 3 on
the ring strikingly modifies the global steric strain E′N,n
,
which can be ascribed to a limitation of the number of
conformations to be “frozen” in the TS, i.e., a change in
the activation entropy;
•The effect of the hybridization sp2/sp3 or the substi-
tution on other positions than position 3 of the ring
depends on the ring size; i.e., no effect of the substituent
type or the hybridization in position 4 of a six-membered
ring whereas the presence of sp2
hybridized atom in the
position 4 or 5 in a seven- or eight-membered ring,
respectively, increases the steric constant r.
However, it is remarkable that eq 9 is sufficiently
versatile and robust to account for the reactivity of 43
alkoxyamines out of 54 based on cyclic nitroxyl frag-
ments. This versatility and robustness merely requires
the generation of 8 new constants (Table 3) for five-,
six-, seven- and eight-membered rings. Such an equation
should play a pivotal role for designing more new
efficient alkoxyamines.
Acknowledgment. S.R.A.M. is deeply grateful for
the guidance of and fruitful discussions with Prof.
Fischer during the 3 years spent in the Physical
Chemistry Institute at the University of Zurich. The
authors thank the Swiss National Foundation for

Macromolecules, Vol. 38, No. 24, 2005
C-ON Bond Homolysis Rate 9983
(36) Marque, S.; Fischer, H.; Baier, E.; Studer, A. J. Org. Chem.
2001, 66, 1146-1156.
(37) Cameron, N. R.; Reid, A. J. Macromolecules 2002, 35, 9890-
9895.
Scientific Research. S.R.A.M. thanks the University of
Zu¨rich and University of Provence for their financial
support.
(38) Marque, S.; Sobek, J.; Fischer, H.; Kramer, A.; Nesvadba,
P.; Wunderlich, W. Macromolecules 2003, 36, 3440-3442.
(39) Miura, Y.; Nakamura, N.; Taniguchi, I.; Ichikawa, A. Polymer
2003, 44, 3461-3467.
(40) Schulte, T.; Studer, A. Macromolecules 2003, 36, 3078-3084.
(41) Cameron, N. R.; Bacon, C. A.; Reid, A. J. ACS Symposium
Series 854; American Chemical Society: Washington, DC,
2003; pp 452-465.
(42) Miura, Y.; Ichikawa, A.; Taniguchi, H. Polymer 2003, 44,
5187-5194.
(43) Miura, Y.; Nakamura, N.; Taniguchi, I.; Ichikawa, A. Polymer
2003, 44, 3461-3467.
(44) Wetter, C.; Gierlich, J.; Knoop, C. A.; Mu¨ller, C.; Schulte, T.;
Studer, A. Chem.sEur. J. 2004, 10, 1156-1166.
(45) Knoop, C.; Studer, A. J. Am. Chem. Soc. 2003, 125, 16327-
16333.
(46) Marque, S. J. Org. Chem. 2003, 68, 7582-7590.
(47) Exner, O. J. Phys. Org. Chem. 1999, 12, 265-274.
(48) Charton, M. J. Phys. Org. Chem. 1999, 12, 275-282.
(49) Galkin, V. I. J. Phys. Org. Chem. 1999, 12, 283-288.
(50) Exner, O.; Charton, M.; Galkin, V. J. Phys. Org. Chem. 1999,
12, 289.
Supporting Information Available: Figure 1SI, showing
a plot of log kd vs eq 1 for alkoxyamines 4′-45. This material
is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(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.
(3) Hawker, C. J. Acc. Chem. Res. 1997, 30, 373-382.
(4) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. Rev. 2001,
101, 3661-3688.
(5) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925-3927.
(6) Fischer, H. Macromolecules 1997, 30, 5666-5672.
(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.
(9) Fischer, H. ACS Symposium Series 854; American Chemical
Society: Washington, DC, 2003; Chapter 2, p 10.
(10) Fischer, H. J. Polym. Sci., Part A: Polym. Chem. 1999, 37,
1885-1901.
(11) Fischer, H.; Souaille, M. Chimia 2001, 55, 109-113.
(12) Fischer, H.; Souaille, M. Macromol. Symp. 2001, 174, 231-
240.
(13) Fukuda, T.; Goto, A.; Ohno, K. Macromol. Rapid Commun.
2000, 21, 151-165.
(14) Fukuda, Y.; Goto, A.; Tsujii, Y. In Handbook of Radical
Polymerization; Matyjaszewski, K., Davis, P. T., Eds.; Wiley-
Interscience: New York, 2002; p 407.
(51) Authors cited in refs 47-50 agree on the universality of the
electrical effect, and that the most common σ*, σI, σF, and σl
Hammett constants have the same meaning.
(52) Taft, R. W. In Steric Effects in Organic Chemistry; Newman,
M. S., Ed.; Wiley: New York, 1956; Chapter 13.
(53) Fujita, T.; Takayama, C.; Nakajima, M. J. Org. Chem. 1973,
38, 1623-1631.
(54) Molawi, K.; Schulte, T.; O.; S. K.; Wetter, C.; Studer, A.
Chem.sEur. J. 2005, 11, 2335-2350.
(55) For 11: Hawker, C. J.; Barclay, G. G.; Dao, J. J. Am. Chem.
Soc. 1996, 118, 11467-11471.
(15) Goto, A.; Fukuda, T. Prog. Polym. Sci. 2004, 29, 329-385.
(16) Bertin, D.; Chauvin, F.; Marque, S.; Tordo, P. Macromolecules
2002, 35, 3790-3791.
(56) For 18: Zink, M.-O.; Kramer, A.; Nesvadba, P. Macromol-
ecules 2000, 33, 8106-8108.
(57) For 20, 26, and 44: Nesvadba, P.; Bugnon, L.; Sift, R. J.
Polym. Sci., Part A: Polym. Chem. 2004, 42, 3332-3341.
(58) For 25, 33, and 34: Nesvadba, P.; Zink, M.-O.; Kramer, A.
DE 19909767 1999; Chem. Abstr. 1999, 131, 229170.
(59) For 32: Nesvadba, P.; Zink, M.-O.; Kramer, A. DE 10113209,
2001; Chem. Abstr. 2001, 135, 273355.
(17) Guerret, O.; Couturier, J. L.; Chauvin, F.; El-Bouazzy, H.;
Bertin, D.; Gigmes, D.; Marque, S.; Fischer, A.; Tordo, P. ACS
Symp. Ser. 2003, 854, 412-423.
(18) Fukuda, T.; Goto, A. Macromol. Rapid Commun. 1997, 18,
683-688.
(19) Bon, S. A. F.; Chambard, G.; German, A. L. Macromolecules
1999, 32, 8269-8276.
(60) Perrin, D. D.; Armarego, W. L. F. In Purification of Labora-
tory Chemicals, 3rd ed.; Pergammon Press: Oxford, U.K.,
1988.
(20) Goto, A.; Fukuda, T. Macromol. Chem. Phys.; 2000, 201,
2138-2142.
(61) For 3a, the supplementary crystallographic data (CCDC-
281943) for this paper have been deposited with the Cam-
bridge Crystallographic Data Centre. They can be obtained,
on request, from the Director, Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge, CB2 1EZ, U.K.
These data may also be obtained free of charge from the
Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
(62) Kirner, H.-J.; Schwarzenbach, F.; Van Der Schaaf, P. A.;
Hafner, A.; Rast, V.; Frey, M.; Nesvadba, P.; Rist, G. Adv.
Synth. Catal. 2004, 346, 554-560.
(63) Goto, A.; Kwak, Y.; Yoshinobu, T.; Tsujii, Y.; Sugiura, Y.;
Fukuda, T. Macromolecules 2002, 35, 3520-3525.
(64) An increment of +14.3 kJ/mol is given by Marque et al. in
ref 24, but recent unpublished work of Marque suggests a
value of +18.0 kJ/mol, which is used.
(21) Ananchenko, G. S.; Fischer, H. J. Polym. Sci.: Part A: Polym.
Chem. 2001, 39, 3604-3621.
(22) Ananchenko, G. S.; Souaille, M.; Fischer, H.; Le Mercier, C.;
Tordo, P. J. Polym. Sci.: Part A: Polym. Chem. 2002, 4,
3264-3283.
(23) 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
Symposium Series 768; American Chemical Society: Wash-
ington, DC, 2000; pp 108-122.
(24) Marque, S.; Le Mercier, C.; Tordo, P.; Fischer, H. Macro-
molecules 2000, 33, 4403-4410.
(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) Acerbis, S.; Beaudoin, E.; Bertin, D.; Gigmes, D.; Marque,
S.; Tordo, P. Macromol. Chem. Phys. 2004, 205, 973-978.
(27) Bertin, D.; Gigmes, D.; Le Mercier, C.; Marque, S. R. A.;
Tordo, P. J. Org. Chem. 2004, 69, 4925-4930.
(65) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119-251.
(66) Charton, M. Prog. Phys. Org. Chem. 1987, 16, 287-315.
(67) Charton, M. In Advances in Quantitatives Structure-Property
Relationship; Charton, M., Ed.; Jai Press Inc.: Greenwich,
CT, 1996; Vol. 1, pp 171-219.
(28) Bertin, D.; Gigmes, D.; Marque, S. R. A.; Maurin, R.; Tordo,
P. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 3504-
3515.
(29) Moad, G.; Rizzardo, E. Macromolecules 1995, 28, 8722-8728.
(30) Stipa, P.; Greci, L.; Carloni, P.; Damiani, E. Polym. Degrad.
Stab. 1997, 55, 323-327.
(31) Matyjaszewski, K.; Woodworth, B. E.; Zhang, X.; Gaynor, S.
G.; Metzner, Z. Macromolecules 1998, 31, 5955-5957.
(32) Skene, W. G.; Belt, S. T.; Connolly, T. J.; Hahn, P.; Scaiano,
J. C. Macromolecules 1998, 31, 9103-9105.
(33) Ciriano, M. V.; Korth, H.-G.; van Scheppingen, W. B.; Mulder,
P. J. Am. Chem. Soc. 1999, 121, 6375-6381.
(68) Charton, M. Stud. Org. Chem. 1991, 42, 629-692.
(69) Unger, S. H.; Hansch, C. Prog. Phys. Org. Chem. 1976, 12,
91-118.
(70) Bacon, C. A.; Cameron, N. R.; Reid, A. J. Macromol. Chem.
Phys. 2003, 204, 1923-1932.
(71) Cameron et al.41,70 measured TEMPO-PhEt and observed a
value 2.7 kJ/mol smaller than the one estimated by Marque
et al.24 and other authors.32,39,63.
(72) A very close polar effect is expected for MeNR1R2 and PhCH2-
NR1R2 groups; see refs 65-67.
(34) Skene, W. G.; Scaiano, J. C.; Listigover, N. A.; Kazmaier, P.
M.; Georges, M. K. Macromolecules 2000, 33, 5065-5072.
(35) Bertin, D.; Gigmes, D.; Marque, S.; Milardo, S.; Peri, J.; Tordo,
P. Collect. Czech. Chem. Commun. 2004, 69, 2223-2238.
(73) Measures of kd on new five-membered ring nitroxyl fragment
based alkoxyamines showed that the presence of substituents
in the positions 3 and 4 of the ring increased markedly the
values of kd, and the new data obeyed eq 1 when r(C5) is

9984 Fischer et al.
Macromolecules, Vol. 38, No. 24, 2005
decreased by 0.11. Marque, S. R. A.; Bagryanskaya, E.;
Kirilyuk, I. A. Personal communication.
(74) Values estimated with r(C5) (coordinates are 0.99 for the x
axis and -4.15 for the y axis for 4′, coordinates are 0.99 for
the x axis and -3.88 for the y axis for 6′) are shown as
triangles in Figure 1SI in the Supporting Information, and
their Es values are italicized in Table 2.
(75) Derwan, P.; Aldabbagh, F.; Zetterlund, P. B.; Yamada, B. J.
Polym. Sci.: Part A: Polym. Chem. 2002, 41, 327-334.
(76) Charton, M. Top. Curr. Chem. 1983, 114, 57-91.
(77) r(CH2OH) is given by Es ) 1.052 - 2.096υ; see: Gallo, R.
Prog. Phys. Org. Chem. 1983, 14, 115-163.
(89) He´berger, K.; Keme´ny, S.; Vidoczy, T. Int. J. Chem. Kinet.
1987, 19, 171-181.
(90) Cameron et al.41,70 estimated values of A far from of 2.4 ×
1014 s-1 for TEMPO-PhEt type alkoxyamines and provided
the rate constants. It can be shown that the values of Ea and
A depend on the choice of the fitting method, either logarith-
mic or exponential fits, and that setting either A or Ea values
equal to those for TEMPO-PhEt given in ref 24, fits are as
good as those obtained with free parameters. All these items
are discussed in more detail in a forthcoming review on the
reactivity of alkoxyamines. Marque, S. R. A.; Gigmes, D.
Manuscript in preparation.
(78) The methyl group can be replaced by an oxo group and then
sp3 hybridized carbon by sp2 hybridized carbon; see ref 46.
(79) Benson, S. W. In Thermochemical Kinetics; John Wiley &
Sons: New York, 1968; pp 55-119.
(91) Dubois, J.-E.; MacPhee, J. A.; Panaye, A. Tetrahedron 1980,
36, 919-928.
(92) The values of r(COOMe) span from 0 to -1.86, and -0.42
gives the best fits. See refs 68 and 77.
(80) A reviewer questioned that five-, six-, seven-, and eight-
membered rings belong to the same family of closely related
compounds.
(93) In the text, it was mentioned that the cis/trans configuration
around the nitroxyl moiety has no significant influence for
such alkoxyamines. However, such configurations might have
an effect on the IHB occurrence in the nitroxyl radical.
(81) Assuming that eqs 1-7 hold and with σI,46 ) -0.02 and Es,46
) -3.45.
(82) Anderson, A. G.; Tocher, D. A.; Corrie, J. E. T.; Lunazzi, L.
J. Am. Chem. Soc. 1993, 115, 3494-3498.
(94) aN,21• ≈ aN,22• as expected for weakly electron-withdrawing
groups such as CH2OSiMe2t-Bu (see Table 4), as shown in
refs 46 and 65. Contrary to what was expected, aN,41• is very
low. Consequently, the influence of the electron-withdrawing
capacity of the CH2OSiMe2t-Bu groups seems to depend on
the conformation around the nitroxyl moiety. It has to be
mentioned that the values of aN for 16•, 22•, and 41• reported
in ref 45 were differents1.530, 1.466, and 1.416 mT,
respectivelysfrom the ones reported in this paper. The new
values were measured from samples kindly provided by Pr.
Studer. The differences observed are due to misprints in ref
45.
(83) Casarini, D.; Lunazzi, L. Anderson, J. E. J. Org. Chem. 1993,
58, 714-717.
(84) Ali, S. A.; Hassan, A.; Wazeer, M. I. M. J. Chem. Soc., Perkin
Trans. 2 1996, 1479-1483.
(85) Raban, M.; Kost, D. Tetrahedron 1984, 40, 3345-3381.
(86) Busfield, W. K.; Jenkins, I. D.; Thang, S. H.; Moad, G.;
Rizzardo, E.; Solomon, D. H. J. Chem. Soc., Chem. Commun.
1985, 1249-1250.
(87) Anderson, J. E.; Casarini, D.; Corrie, J. E. T.; Lunazzi, L. J.
Chem. Soc., Perkin Trans. 2 1993, 1299-1304.
(88) Bertin, D.; Gigmes, D.; Marque, S. R. A.; Tordo, P. Macro-
molecules 2005, 38, 2638-2650.
MA0512612