Reactivity of polyolefins toward cumyloxy radical: Yields and regioselectivity of hydrogen atom transfer

Article abstract of DOI:10.1021/ma402177v

Hydrogen atom abstraction from a series of homopolymers by cumyloxy radicals is examined under solvent-free conditions at temperatures that are relevant to radical-mediated polymer modifications. Abstraction efficiency data acquired for the thermolysis of dicumyl peroxide (DCP) within pure polymer samples establish an order of reactivity: poly(butadiene) (PBD) > poly(ethylene oxide) (PEO) > poly(ethylene) (PE) > poly(propylene) (PP) > poly(isobutylene) (PIB). The regioselectivity of hydrogen transfer from PE, PP, and PIB is assessed by model hydrocarbon experiments involving nitroxyl trapping of the alkyl radicals generated from pentane, 2,4-dimethylpentane, and 2,2,4,4-tetramethylpentane, respectively. Taken together, abstraction efficiency and regioselectivity data are discussed in terms of enthalpic and entropic contributions to H atom transfer rates, with particular emphasis on steric hindrance imposed by methyl substituents on secondary positions within PP and PIB. The utility of polymer oxidizability as a predictive measure of the reactivity of a polymer toward cumyloxyl and vinyltriethoxysilane graft modification is evaluated and discussed.

Full text of DOI:10.1021/ma402177v

Article
pubs.acs.org/Macromolecules
Reactivity of Polyolefins toward Cumyloxy Radical: Yields and
Regioselectivity of Hydrogen Atom Transfer
†
†
‡,§
,†
Graham E. Garrett, Elena Mueller, Derek A. Pratt, and J. Scott Parent*
†
Department of Chemical Engineering, Queen’s University, Kingston, ON Canada K7L 3N6
‡
Department of Chemistry, Queen’s University, Kingston, ON Canada K7L 3N6
§
Department of Chemistry, University of Ottawa, Ottawa, ON Canada K1N 6N5
ABSTRACT: Hydrogen atom abstraction from a series of homopolymers by
cumyloxy radicals is examined under solvent-free conditions at temperatures
that are relevant to radical-mediated polymer modifications. Abstraction
efficiency data acquired for the thermolysis of dicumyl peroxide (DCP) within
pure polymer samples establish an order of reactivity: poly(butadiene) (PBD) >
poly(ethylene oxide) (PEO) > poly(ethylene) (PE) > poly(propylene) (PP) >
poly(isobutylene) (PIB). The regioselectivity of hydrogen transfer from PE, PP,
and PIB is assessed by model hydrocarbon experiments involving nitroxyl
trapping of the alkyl radicals generated from pentane, 2,4-dimethylpentane, and
2,2,4,4-tetramethylpentane, respectively. Taken together, abstraction efficiency
and regioselectivity data are discussed in terms of enthalpic and entropic
contributions to H atom transfer rates, with particular emphasis on steric
hindrance imposed by methyl substituents on secondary positions within PP
and PIB. The utility of polymer oxidizability as a predictive measure of the reactivity of a polymer toward cumyloxyl and
vinyltriethoxysilane graft modification is evaluated and discussed.
INTRODUCTION
Much of what is known about the radical reactivity of
polymers is derived from ambient temperature studies of their
autoxidation. This process involves a chain reaction sequence of
alkyl radical combination with oxygen, and H atom abstraction
by the resulting peroxyl radical from the polymer, as shown
■
Solvent-free, peroxide-initiated polymer modifications are used
widely to improve the physical and/or chemical properties of
commodity plastics and elastomers. Important examples
1
include the cross-linking of ethylene-rich polymers into
15
2
,3
below.
thermoset products, the controlled degradation of propy-
lene-rich polymers to improve their processing character-
•
•
R +O → ROO
4
,5
2
istics, and the graft modification of polyolefins with maleic
6
7
•
•
anhydride and vinyltrialkoxysilanes to produce functional
derivatives that are suitable for a range of adhesive, blend, and
ROO +RH → ROOH + R
8
composite formulations. The introduction of long chain
branches to propylene homopolymers has also attracted recent
interest, owing to the improved extensional viscosity provided
Since the latter reaction is rate determining at ambient
temperatures, the susceptibility of a hydrocarbon to oxidation
is influenced greatly by its reactivity as a H atom donor to
peroxyl intermediates. This is reflected in measurements of
substrate oxidizability, defined as the propagation rate constant
for H atom transfer divided by the square root of the peroxyl
radical termination rate constant. Tabulated values have
demonstrated the importance of reaction enthalpy to the
kinetics of this process, with compounds having low C−H bond
dissociation energies (BDEs) (i.e., allylic groups, benzylic
functionalities, and ethers bearing α-hydrogens) being partic-
9
,10
by branched architectures.
As is common in the polymer
industry, commercial developments have outpaced studies of
fundamental chemistry, and an improved understanding of
high-temperature radical addition and atom transfer processes
is needed to support further advances in the field.
To a great extent, the amenability of a polymer with respect
to these technologies is dictated by its reactivity with respect to
11
12
H atom transfer. Cross-linking and controlled degradation
16
ularly susceptible. Limited amounts of oxidizability data on
polymeric systems are consistent with small molecule results,
with oxidizability values for cis-1,4-poly(isoprene) being greater
than those recorded for polyolefins and acrylate-derived
processes are initiated by H atom transfer from the polymer to
peroxide-derived radical intermediates, with ensuing macro-
radical combination and/or fragmentation producing the
intended changes to molecular weight distributions. Monomer
grafting processes similarly require polymer macroradical
generation but also involve H atom transfer between the
polymer and monomer-derived radical adducts to support a
Received: October 22, 2013
Revised: December 13, 2013
Published: January 10, 2014
1
3,14
chain propagation sequence.
©
2014 American Chemical Society
544
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
1
7
polymers, owing to the low BDE of the allylic functionality,
Where a polymer presents multiple H atom donors (e.g.,
polypropylene and polyisobutylene), knowledge of the relative
reactivity of primary, secondary, and tertiary positions toward
H atom transfer is required. These regioselectivity studies are
usually performed on model hydrocarbons, owing to the
difficulty of characterizing polymer reaction products unambig-
and poly(propylene) is more easily oxidized than poly-
ethylene), 20d,2u1e to the reactivity of its tertiary H atom
1
8,19
(
•
toward ROO .
For oxidizability to be a useful proxy for the polymer
modifications described above, H atom transfer to peroxyl
radicals must correlate with H atom transfer to alkoxy and alkyl
radical intermediates. However, significant differences in the
reactivity of peroxyl and alkoxyl radicals are widely
3
1,32
uously.
Electron spin resonance analysis of alkyl radicals or
33,34
their spin-trapped derivatives has been used,
as has the
quantification of alkoxyamine products generated by nitroxyl
2
2,23
•
35
recognized.
Indeed, H atom transfer to t-BuO can be
trapping. Solomon and co-workers used the latter technique
dominated by entropic effects, with transition states possessing
to gain insight into PP activation by studying the products
24
relatively small enthalpies of activation. Factors affecting the
entropy of activation, including steric effects, can be controlling
in these systems, making simple assumptions concerning the
reactivity of polymeric C−H donors unreliable. Therefore,
oxidizability may be a poor indicator of polymer reactivity, and
a different measure may be required.
A complete understanding of the H atom transfers involved
in polyolefin graft modification requires knowledge of
abstraction by alkoxyl-, methyl-, and monomer-derived radicals
from every distinguishable H atom donor presented by a
polymer. This is a formidable challenge. However, considerable
insight can be gained by examining the byproducts of dicumyl
peroxide (DCP) thermolysis (Scheme 1). Since the rate of
derived from H atom transfer from 2,4-dimethylpentane to tert-
36
butoxyl at 60 °C. We have extended their approach to include
model compounds for PE and PIB, using cumyloxyl at
temperatures that are relevant to industrial practice.
A secondary objective of this work was to explore the
relationship between AE measurements of H atom donation
and the efficiency of polymer grafting modifications. The
functionalization of a polymer with single grafts of maleic
anhydride and vinyltrimethoxysilane (VTMS) is formally a C−
H addition from the polymer to a CC bond. Initiation occurs
by H atom abstraction from the polymer to peroxide-derived
alkoxyl radicals to produce a macroradical, whose propagation
involves addition to an alkene as well as H atom abstraction.
Therefore, polymers that are poor H atom donors are expected
to be more difficult to functionalize by radical chemistry,
providing lower graft yields and a smaller proportion of single
graft adducts. Insight into the relationship between H atom
abstraction efficiency and graft modification efficiency is
provided through comparisons of AE data with independent
measurements of VTMS addition yields for a range of saturated
polymer substrates.
Scheme 1
EXPERIMENTAL SECTION
Instrumentation and Analysis. NMR spectra were recorded on a
■
1
13
Bruker spectrometer in CDCl at 400 and 100 MHz for H and
C
3
measurements, respectively. High-resolution mass spectra were
acquired with a QStar XL QqTOF instrument. Reverse-phase HPLC
was conducted using a Waters 600E system controller employing a
model 600 pump, a model 2487 UV detector, and an XBridgeTM
semipreparative C18 column (10 by 150 mm) with 5 μm particle size.
GC-MS analysis was conducted using an Agilent Technologies 6850
Network GC system coupled with a 5975C VL MS detector. The
instrument was configured with an Agilent Technologies 19091S-433e
HP-5MS column with an inlet temperature of 250 °C, pressure of 78
kPa, and flow rate of 52 mL/min. The initial temperature was 70 °C
and was increased at a rate of 6 °C/min to 280 °C, where it was held
for 5 min. Abstraction efficiency measurements derived from cumyl
alcohol and acetophenone yields were determined using a Hewlett-
Packard 5890 series II gas chromatograph equipped with a Supelco
SPB-1 microbore column using 2 mL/min of helium as carrier gas.
Injector and detector temperatures were held at 225 and 300 °C,
respectively, with the oven temperature starting at 40 °C for 6 min,
ramping to 150 °C at 10 °C/min, ramping to 280 °C at 12 °C/min,
and holding for 15 min. FID calibration was accomplished using
authentic samples.
cumyloxyl radical fragmentation in nonpolar media is solely a
25,26
function of temperature,
the relative amounts of cumyl
alcohol and acetophenone generated within a hydrocarbon
substrate provide a simple measure of the latter’s intrinsic
27
reactivity toward H atom donation. This approach has been
applied to H atom abstraction by cumyloxyl from a range of
2
8,29
saturated hydrocarbons and alkylated aromatics
as well as
H atom donation to tert-butoxyl by a limited range of
3
0
polymers.
In spite of the central role of H atom transfer in polymer
modification chemistry, comprehensive data for the solvent-free
activation of polymers are unavailable. The primary aim of this
work was to quantify and rationalize the H atom abstraction
efficiency (AE = [cumyl alcohol]/([cumyl alcohol] +
Materials. Dicumyl peroxide (DCP, 98%), 2,4-dimethylpentane
DMP, 99%), hexadecane (97%), pentane (99%), and vinyltrimethox-
(
[acetophenone])) of DCP acting on a range of commercial
ysilane (VTMS, 95%) from Sigma-Aldrich and 2,2,4,4-tetramethyl-
pentane (TMP, 98%) from Alfa Aesar were used as received. Di-tert-
butyl hyponitrite was prepared by the esterification of sodium
homopolymers at temperatures that are relevant to industrial
practice. The resulting data, acquired for purified samples of cis-
1
,4-poly(butadiene) (PBD), poly(ethylene oxide) (PEO), high-
hyponitrite by tert-butyl bromide in the presence of ZnBr according
2
density poly(ethylene) (PE), poly(propylene) (PP), and
poly(isobutylene) (PIB), provide a consistent reference set
for gauging the responsiveness of these materials toward
chemical modification.
37
to the method of Mendenhall. Polyisobutylene (PIB, M = 85 000)
w
and atactic-polypropylene (PP, M = 10 000) from Scientific Polymer
w
Products were hydrogenated prior to use by treatment of polymer
dissolved in hexane with platinum supported on carbon at 20 bar of H2
5
45
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
gas, 100 °C for 50 h, after which the polymer was recovered by
precipitation from acetone and dried under vacuum. Polybutadiene
145.42; 67.18. HRMS (EI) m/z calcd for C H NO: 289.2406; found
19
31
289.2412. HPLC: t = 35.3 min.
R
(
PBD, Mn = 5000, Scientific Polymer Products) was purified by
2-(2,4-Dimethylpentan-3-yloxy)-1,1,3,3-tetramethylisoindoline
1
dissolution/precipitation (hexanes/methanol) and dried under vac-
uum. Poly(ethylene glycol) monomethyl ether (PEO, M = 5000, Alfa
Aesar) purified by dissolution/precipitation (THF/hexane) and dried
under vacuum. Polyethylene (HDPE, M = 1800, Scientific Polymer
Products) (13.23 g) was dissolved toluene at 90 °C and precipitated
from acetone before drying under vacuum.
(2s). H NMR (CDCl
, 400 MHz): δ ppm = 7.22 (dd, 2H, J = 5.5, 3.2
3
Hz); 7.09 (dd, 2H, J = 5.5, 3.2 Hz); 3.47 (s, 1H); 2.26 (m, 2H); 1.51
n
(
7
9
br s, 6H); 1.40 (br s, 6H); 1.04 (d, 6H, J = 7.0 Hz); 1.00 (d, 6H, J =
.0 Hz). 13C NMR (CDCl , 100 MHz): δ ppm = 127.04; 121.37;
3
n
0.21; 28.13; 30.24; 25.62; 20.00; 19.67; 139.39; 67.18. HRMS (EI)
m/z calcd for C H NO: 289.2406; found 289.2416. HPLC: t = 38.8
19 31
R
min.
1
,1,3,3-Tetramethylisoindolin-2-oxyl (TMIO). 10% Pd/C (0.38 g)
2
-(2,4-Dimethylpentan-2-yloxy)-1,1,3,3-tetramethylisoindoline
was placed in a 100 mL round-bottom flask, evacuated, and filled with
1
38
(2t). H NMR (CDCl , 400 MHz): δ ppm = 7.22 (dd, 2H, 5.6, 3.2
3
argon. 2-Benzyl-1,1,3,3-tetramethylisoindoline (3.56 g, 13.4 mmol)
was dissolved in methanol (45 mL) and added dropwise prior to the
addition of ammonium formate (5.23 g, 82.9 mmol). The mixture was
heated to reflux for 12 h, cooled, and filtered through Celite prior to
concentrating under vacuum to yield crude product (4.26 g), which
was oxidized directly and recrystallized as described by Griffiths et al.
Hz); 7.08 (dd, 2H, 5.6, 3.2 Hz) 1.85 (m, 1H); 1.52 (d, 2H, J = 5.6
Hz); 1.48 (s, 6H); 1.32 (s, 6H); 1.29 (s, 6H); 0.98 (d, 6H, J = 6.7).
13
C NMR (CDCl , 100 MHz): δ ppm =121.69; 126.99; 24.57; 30.70;
3
5
2.09; 27.51; 25.95; 30.70; 25.15; 145.68; 68.09; 79.13;. HRMS (EI)
m/z calcd for C H NO: 289.2406; found 289.2417. HPLC: t = 43.4
19 31
R
38
min.
to give TMIO (1.85 g, 9.7 mmol).
Preparation of 2,2,4,4-Tetramethylpentane-Derived Alkoxy-
amines 3p and 3s. The reaction and product isolation were carried
out using 2,2,4,4-tetramethylpentane according to the procedure
described above for 2,4-dimethylpentane, except for the use of a 90/10
Preparation of Pentane-Derived Alkoxamines 1p, 1s, and
s′. Di-tert-butyl hyponitrite (1.0 mmol) and 1,1,3,3-tetramethyli-
1
soindolin-2-oxyl (2.2 mmol) in n-pentane (60 mmol) were
deoxygenated by freeze−thaw degassing and heated under argon to
3
concentrated using a rotary evaporator, diluted in methanol to a
concentration of 5 mg/100 μL, and separated by HPLC (85/15
MeOH/H O, flow rate = 4 mL/min, UV = 270 nm). Fractions were
combined and concentrated under reduced pressure until a white
precipitate began to form, but not until dryness. Concentrated solutions
were extracted with dichloromethane, dried (MgSO ), and filtered.
The organic layer was concentrated, diluted with methanol, and
extracted once with an equal volume of hexanes. The hexanes were
removed under reduced pressure, leaving alkoxyamines 1p and a
mixture of 1s and 1s′ as clear viscous oils.
39
MeOH/H O HPLC eluent mixture.
2
0 °C for 11 days (4 initiator half-lives). The resulting mixture was
1
,1,3,3-Tetramethyl-2-(2,2,4,4-tetramethylpentan-3-yloxy)-
1
isoindoline (3p). H NMR (CDCl , 400 MHz): δ ppm = 7.23 (dd,
3
2
H, J = 5.5, 3.2 Hz); 7.10 (dd, 2H, J = 5.5, 3.2); 3.67 (s, 2H); 1.45 (br
2
13
s, 12H); 1.42 (s, 2H); 1.12 (s, 6H); 1.03 (s, 9H). C NMR (CDCl3,
1
1
3
00 MHz): δ ppm = 121.54; 127.22; 88.21; 52.27; 32.24; 26.78;
45.65; 67.51; 36.99; 29.86. HRMS (EI) m/z calcd for C H NO:
1
9
31
4
17.2719; found 317.2731. HPLC: t = 20.7 min.
R
1
,1,3,3-Tetramethyl-2-(2,2,4,4-tetramethylpentyloxy)isoindoline
1
(
3s). H NMR (CDCl , 400 MHz): δ ppm = 7.22 (dd, 2H, J = 5.6, 3.2
3
Hz); 7.07 (dd, 2H, J = 5.6, 3.2 Hz); 3.71 (s, 1H); 1.64 (br s, 6H); 1.54
13
(
1
br s, 6H); 1.22 (s, 18H). C NMR (CDCl , 100 MHz): δ ppm =
3
1
,1,3,3-Tetramethyl-2-(pentyloxy)isoindoline (1p). 1H NMR
27.21; 121.24; 93.04; 31.54; 146.07; 69.48; 38.83. HRMS (EI) m/z
(
(
CDCl , 400 MHz): δ ppm = 7.22 (dd, 2H, J = 5.5, 3.2 Hz); 7.09
3
calcd for C H NO: 317.2719; found 317.2727. HPLC: t = 25.7 min.
19
31
R
dd, 2H, J = 5.5, 3.2 Hz); 3.91 (t, 2H, 6.6 Hz); 1.64 (m, 2H); 1.42 (br
Regioselectivity Measurements. A 10 mL stainless steel vessel
was charged with dicumyl peroxide (0.05 mmol), 1,1,3,3-tetramethy-
lisoindolin-2-oxyl (0.1 mmol), and pentane (6 mmol) and degassed by
pressurizing and releasing 24 bar of argon. The vessel was then sealed
under 24 bar of argon and heated to 160 °C for 5 min (1 initiator half-
life). Residual pentane was removed by rotary evaporation, and the
products were diluted with MeOH to a concentration of 5 mg/100 μL
prior to GC-MS analysis of relative amounts of 1s and 1s′. Reverse
phase HPLC (85:15 methanol/water; 4 mL/min) fractionation of the
sample followed by UV−vis peak integration provided the relative
yields of 1p and 1s + 1s′ under the assumption that the extinction
13
s, 12H); 1.26 (m, 4H); 0.98−0.92 (m, 3H). C NMR (CDCl , 100
MHz): δ ppm = 127.08; 121.44; 99.97; 34.65; 29.05, 28.68; 25.27,
2
2
3
0.68; 11.41. HRMS (EI) m/z calcd for C H NO: 261.2093; found
17 27
61.2088. HPLC: t = 20.7 min.
R
1
,1,3,3-Tetramethyl-2-(pentan-2-yloxy)isoindoline and 1,1,3,3-
Tetramethyl-2-(pentan-3-yloxy)isoindoline (1s and 1s′). Although
these structural isomers were inseparable by HPLC, the isomers were
separated from the primary abstraction products and characterized
partially by NMR. The CH protons adjacent to the alkoxyamines of
the two isomers were identified by their splitting patterns based on the
symmetry of 1s′ versus 1s (3.91 and 3.70 ppm, respectively). These
protons integrated at a 1:3.15 ratio of 1s′:1s. To determine if this ratio
was artificially caused by the HPLC product separation, the mixture of
secondary products, as well as the isolated 1p, was analyzed by GC/
MS, revealing a 1:2.96 ratio of 1s′:1s, agreeing well with the NMR
results. By comparison, a sample of the crude reaction mixture gave a
ratio of secondary products (1s′:1s) of 1:2.39. H NMR (CDCl , 400
MHz): δ ppm = 7.22 (m, 8.25H) 7.09 (m, 8.30H); 3.91 (h, 3.15H, J =
5
2
40
coefficient of each alkoxyamine at 270 nm is identical. The overall
yield of 1p:1s:1s′ was 12:56:31.
Alkoxyamines derived from 2,4-dimethylpentane were produced as
described for pentane using reaction times of 5, 8, and 10 min. Product
concentrations at each reaction time were quantified by integration of
fully resolved HPLC UV−vis signals. The observed distributions of
2p:2s:2t were 31:7:63 at 5 min, 32:6:62 at 8 min, and 33:5:61 at 10
min, indicating that tertiary alkoxyamine instability was not problem-
atic under the reaction conditions. Extrapolation of these yields to time
zero provided a 2t:2p:2s distribution of 28:8:64.
1
3
.9 Hz); 3.70 (p, 1H, 5.7 Hz); 1.79−1.59 (m, 8.08H); 1.51−1.49 (s,
4.24H); 1.45−1.42 (m, 10.34H); 1.37 and 1.32 (brs, 25.56H); 1.23
(
d, 9.45H); 0.96 (m, 15.66H).
Radical trapping products of 2,2,4,4-tetramethylpentane were
generated as described for pentane, with HPLC UV−vis peak
integration providing a 3p:3s distribution of 99:1.
Abstraction Efficiency Studies. H atom abstraction yields from
small molecules (pentane, 2,4-dimethylpentane, 2,2,4,4-tetramethyl-
pentane, hexadecane) were determined by charging a 10 mL stainless
steel vessel with hydrocarbon (1.0 g) and DCP (0.02 g) and
Preparation of 2,4-Dimethylpentane-Derived Alkoxamines
p, 2s, and 2t. The reaction and product isolation were carried out
2
using 2,4-dimethylpentane using the procedure described above for
pentane, except for a reaction temperature of 55 °C for 20 h (11
initiator half-lives).
2
-(2,4-Dimethylpentyloxy)-1,1,3,3-tetramethylisoindoline (2p).
1
H NMR (CDCl , 400 MHz): δ ppm = 7.22 (dd, 2H, J = 5.5, 3.2
deoxygenating the solution by three cycles of pressurizing with N to
3
2
Hz); 7.09 (dd, 2H, J = 5.5, 3.2 Hz); 3.81 (dd, 1H, J = 8.7, 5.1 Hz);
14 bar, mixing, and releasing. The vessels were then pressured to 14
3
1
6
.69 (dd, 1H, 8.7, 6.9 Hz); 1.88 (m, 1H); 1.71 (m, 1H), 1.43 (br s,
bar of N and placed in an oil bath at 160 °C under constant magnetic
2
2H); 1.33−1.05 (m, 2H); 1.01 (dd, 3H, J = 6.7 Hz); 0.92 (d, 3H, J =
stirring for 38 min. The contents were cooled to room temperature
before injecting directly into the GC instrument to determine cumyl
alcohol and acetophenone concentrations. Abstraction efficiencies
.6 Hz); 0.89 (d, 3H, J = 6.6 Hz). 13C NMR (CDCl , 100 MHz): δ
3
ppm = 127.06; 121.42; 83.08; 31.03; 25.31; 17.94; 23.34; 23.36;
5
46
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
were calculated as [cumy alcohol]/([cumyl alcohol] + [acetophe-
none]). Note that cumyl alcohol dehydration to α-methylstyrene was
not observed under the conditions used.
Abstraction efficiencies for polymeric substrates (PBD, PEO,
HDPE, PP, PIB) were measured by coating the material (0.5 g)
with a small volume of DCP in acetone (∼0.2 mL) to produce a
peroxide concentration of 2 wt %. For ground samples of
semicrystalline polymers (PEO, PE), the solvent was allowed to dry,
and the resulting powder was tumble-mixed. For amorphous materials
(
PP, PBD, PIB), samples were dried and passed through a 2-roll mill
to ensure homogeneity. Each mixture was then charged to an Atlas
Laboratory Mixer at 140, 150, 160, 170, and 180 °C for seven initiator
half-lives (302, 105, 39, 14, and 6 min, respectively). The product was
dissolved in toluene (3 mL) and precipitated from acetone (25 mL).
The resulting solution was further diluted and the contents analyzed
by GC for acetophenone and cumyl alcohol. Note that the pooled
standard deviation of AE data was 0.02.
Figure 1. DCP abstraction efficiency as a function of reaction
temperature.
VTMS Grafting. Small molecule grafting experiments involved
charging a 10 mL stainless steel vessel with hydrocarbon (1 g), VTMS
(
0.05 g), and DCP (0.002g), deoxygenating, and pressurizing as
described above. Samples were maintained under continuous stirring
at 160 °C for 38 min before cooling and sampling for H NMR
analysis. VTMS conversion was measured by integration of resonances
43
butoxyl) compared to that for H atom atom transfer (3.9
1
44
kcal/mol for methoxyl abstraction from isobutane). This
insensitivity has practical implications for a wide range of
polymer modifications, as it indicates that the fate of DCP-
derived radicals is not strongly temperature dependent.
Moad has discussed the importance of alkoxy radical
fragmentation in polyolefin modification chemistry, reviewing
differences in the reactivity of methyl and alkoxy radicals in
both H atom transfer and monomer addition. While cumyloxyl
is relatively electrophilic and demonstrates a preference for H
atom abstraction, the opposite is true for methyl radicals, in
derived from residual vinyl functionality (δ 5.8−6.2 ppm, 3H, −CH
CH ) relative to those generated by methoxy substituents (δ 3.6 ppm,
2
s, 9H, −SiOCH ).
3
Samples for polymer grafting were prepared as described above for
abstraction efficiency measurements, resulting in mixtures containing
polymer (0.5 g), VTMS (0.025 g), and DCP (0.001 g), which were
heated to 160 °C for 25 min. The graft-modified samples were purified
6
1
by dissolution/precipitation and dried prior to analysis by H NMR.
1
H NMR spectra of polyethylene were recorded in deuterated toluene
45
at 70 °C using a Bruker AC-400 spectrometer.
that they are weakly nucleophilic and prefer monomer
4
6,47
addition.
Our data show that reaction temperature does
RESULTS
■
not strongly affect the proportions of methyl and polymer
macroradicals generated by DCP. As such, the initiation
efficiency of processes such as PE cross-linking and PP
degradation is expected to be relatively constant over a wide
temperature range. This has been demonstrated in small
molecule systems, with van Drumpt and Oosterwijk showing
that temperature has little effect on the efficiency of
pentadecane cross-linking by tert-butyl perbenzoate between
130 and 165 °C. Similarly, ESR studies of DCP-initiated
macroradical generation from isotactic-PP found no influence
of temperature on initiation efficiency. This consistency in
macroradical yield may, in part, explain why vinyltriethoxysilane
Abstraction Efficiency. The fragmentation of cumyloxyl
4
1
radical is a commonly used radical clock method for
measuring the kinetics of the reaction of cumyloxyl radicals
and small-molecule H atom donors. The approach involves the
determination of the ratio of cumyl alcohol and acetophenone
products observed when DCP is decomposed in the presence
of varying concentrations of an H atom donor. Since the
product ratio depends on the ratio of rate constants (k /k ) and
48
a
d
the H atom donor concentration in solution (Scheme 1), with a
known fragmentation rate constant, the rate constant for the H
atom transfer reaction can be readily determined from a plot of
the product ratios as a function of H atom donor concentration.
Since molar concentrations in neat polymer systems are
ambiguous, we present our data in terms of abstraction
efficiency (AE), defined as the fraction of cumyloxyl radicals
that abstract from the hydrocarbon as opposed to fragmenta-
tion to acetophenone and methyl radicals. AE values measured
for DCP thermolysis in PBD, PEO, PE, PP, and PIB are plotted
as a function of temperature in Figure 1.
The limited amount of literature pertaining to the polymers
of interest compares favorably with the results of this work.
Rowe and Thomas analyzed the byproducts of DCP
thermolysis in poly(ethylene glycol), generating an AE value
of 0.72 at 140 °C, which corresponds to our PEO result of
49
50
addition to PE is insensitive to reaction temperature.
The data plotted in Figure 1 reveal significant differences in
H atom atom donation activity, with AE values declining in the
order PBD > PEO > PE > PP > PIB. The drop observed for the
first three homopolymers parallels the thermodynamic stability
of the resulting secondary radicals. Consider the reported BDE
values for H-3-but-1-enyl (345 kJ/mol) versus H-2-tetrahy-
51
drofuranyl (385 kJ/mol) versus H-cyclohexyl (400 kJ/mol).
Given the BDE of tertiary alcohols (444.9 ± 2.8 kJ/mol for t-
52
BuOH), it is clear that abstraction from all three polymers is
exothermic. Moreover, quantum chemical calculations have
shown that H atom abstraction from simple alkanes by alkoxy
3
4,53
radicals is entropically favorable,
transfer is exergonic.
Since H atom transfer reactions between alkoxyl radicals and
hydrocarbons are highly exergonic, a conventional Evans−
Polanyi correlation of the relevant kinetic and thermodynamic
data (log(k ) = αΔH + constant; ΔH = BDE(C−H) −
BDE(O−H)) is expected to produce a small proportionality
constant α, consistent with an early transition state. Data
and as such, H atom
4
2
0
0
.74. Chodak and Bakos report AE values for cyclohexane of
28
.62 at 145 °C, 0.53 at 160 °C, and 0.49 at 170 °C, while
Dannenberg et al. recorded AE values for mineral oil of 0.69 at
35 °C and 0.55 at 180 °C. These data are comparable to the
1
o
o
PE measurements presented in Figure 1, showing a decline with
increasing temperature that results from a higher activation
energy for cumyloxyl fragmentation (14.6 kcal/mol for tert-
a
5
47
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
published by Finn et al. have demonstrated that, even though
the kinetic reactivity of tert-butoxyl can be dominated by
entropy effects, H atom abstraction from hydrocarbons obeys a
linear Evans−Polanyi relationship with α = 0.13. This
correlation appears to apply to the cumyloxyl system of
present interest, with resonance stabilization of PBD- and PEO-
derived radicals contributing to a heightened kinetic reactivity.
With PBD, PEO, and PE presenting little difference in terms of
steric effects, AE values mirror trends in bond dissociation
energy.
Scheme 2
As described above, the higher oxidizability of PP, coupled
with the presence of tertiary sites of low BDE, suggests that this
material could be more reactive toward cumyloxyl than PE.
However, our data show the opposite, with AE values for PE
proving to be 1.5 times those of PP between 140 and 180 °C.
This result is consistent with low-temperature model
compound studies. Dokolas et al. extended product distribution
data for reactions of tert-butoxyl with 2,4-dimethylpentane and
Table 1. Abstraction Efficiency and Vinyltrimethoxysilane
VTMS) Graft Conversions for Saturated Polymers and
Model Hydrocarbons
(
c
3
-methylpentane at 60 °C to estimate the reactivity of PP and
abstraction
a
VTMS graft
b
peroxide yield
(mol/mol)
54
linear low-density polyethylene (LLDPE), respectively. Their
extrapolations suggest that LLDPE containing 5% comonomer
can provide AE values 1.3 times those of PP. Furthermore,
Camara et al. conducted electron paramagnetic resonance
experiments of H atom atom abstraction by tert-butoxyl at 25
R−H
efficiency
conversion
poly(ethylene oxide)
0.69
1.00
21.6
(
PEO)
polyethylene (HDPE)
polypropylene (PP)
polyisobutylene (PIB)
hexadecane
0.56
0.37
0.17
0.55
0.45
0.39
0.78
0.23
0.10
0.73
0.62
0.22
16.9
5.0
2.2
°
C from a series of model hydrocarbons, generating
−3
15.8
13.4
4.8
bimolecular rate constants of 7.1 × 10 L/(mol s) for decane
and 1.7 × 10−3 L/(mol s) for 2,4-dimethylpentane.
34
pentane
2
,4-dimethylpentane
DMP)
PIB is similar to PP, in that it undergoes chain scission when
treated with low concentrations of dialkyl peroxides at standard
(
2
,2,4,4-
tetramethylpentane
(TMP)
0.18
0.12
2.6
5
5,56
polymer processing temperatures.
However, relatively large
amounts of peroxide are required to bring about significant
molecular weight reductions, and the oxidative stability of PIB
is well established. This lack of radical reactivity is consistent
with the low AE values recorded in this study and is likely the
result of the steric crowding of secondary positions by adjacent
quaternary centers. Unfortunately, the presence of multiple H
atom donors in PIB complicates attempts to interpret AE data.
Our analysis of the regioselectivity of H atom abstraction from
PE, PP, and PIB follows a brief exploration of the relationship
between cumyloxyl radical abstraction efficiencies and polymer
graft modification yields.
Graft Modification. The chemical modification of saturated
polymers by grafting to select monomers is widely practiced,
with an ideal process operating with high kinetic chain length
such that a small amount of peroxide can convert all monomer
into polymer-bound functionality. This requires efficient
macroradical generation by initiator-derived radicals as well as
a
b
7.2 μmol DCP/g; 160 °C; 30 min. 7.2 μmol DCP/g; 320 μmol
c
VTMS/g; 160 °C; 30 min. Peroxide yield = mol VTMS converted/
mol cumyloxyl.
of DCP initiation, and peroxide yields are the moles of VTMS
•
consumed per mole of radicals (cumyloxyl and CH ) derived
3
from DCP thermolysis. Peroxide yields are crude indicators of
the kinetic chain length of VTMS addition, since the initiation
efficiency of graft modification is unknown, and the
concentration of VTMS dropped substantially over the course
of several of these batch reactions. Note that the complete
conversion of 320 μmol VTMS/g using 7.4 μmol DCP/g
corresponds to a peroxide yield of 21.6 mol/mol. VTMS
addition to PEO afforded this maximum value, meaning that
most monomer was consumed through the chain propagation
sequence illustrated in Scheme 2. The relatively low reactivity
of PIB with respect to VTMS addition is reflected by a peroxide
yield of 2.2.
a fast sequence of monomer addition (k ) and H atom atom
add
transfer (k ) to the resulting monomer-derived radical (Scheme
tr
6
,13
2).
To assess the amenability of saturated polymers toward
The correlation between peroxide yield and abstraction
efficiency for saturated hydrocarbons is particularly interesting
(Figure 2). Abstraction efficiency is a specific measurement of
H atom atom donation to cumyloxyl radicals, whereas peroxide
yield relates, in part, to the propagation and termination of
resulting alkyl macroradicals. It is reasonable to assume that
initiation efficiencythe fraction of peroxide-derived radicals
that initiate hydrocarbon addition to VTMSwill improve
with increasing AE. However, the functional dependence
between vinylsilane conversion and initiation efficiency is not
consistent with the observed dependence of conversion on AE.
We have previously demonstrated that vinylsilane conversion
scales with the square root of the product of initiation efficiency
graft modification, we have measured grafting yields for a series
of batch reactions. Vinyltrialkoxysilanes are particularly well
suited to fundamental study, since they do not homopolymerize
5
7,58
readily,
preferring H atom atom abstraction to yield single
graft units versus repeated monomer addition to produce
5
9,60
oligomeric side chains.
Furthermore, they are soluble in
nonpolar hydrocarbons and polyolefins, thereby avoiding
complications associated with phase-partitioning of monomer
and initiator.
In Table 1 is a list of DCP abstraction efficiency and VTMS
conversion data recorded for a range of polymers and small
molecules. AE values are defined as above, while VTMS
conversions are the fraction of monomer consumed as a result
5
0,61
and peroxide concentration,
whereas the data presented in
5
48
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
64
in this work are quite robust, while the stability of the tertiary
alkoxyamine derived from DMP was examined carrying out the
trapping reaction at different peroxide half-lives and verifying
that that the alkoxyamine distribution remained constant.
A comparative analysis of PE, PP, and PIB requires an
account of overall reactivity of the model alkane AE coupled
with knowledge of regioselectivity. Taken together, these
measurements reflect the absolute reactivity of the primary,
secondary and tertiary sites in each substrate (Scheme 3). Of
Scheme 3. Alkyl Radical Trapping Yields (160 °C; 27 mmol
DCP/g; 54 mmol TMIO/g)
Figure 2. Peroxide yield of VTMS grafting versus AE ([DCP] = 0.2 wt
%, [VTMS] = 5 wt %, 160 °C, 38 min).
Table 1 show VTMS conversions increasing monotonically
with AE.
It is reasonable to assume that AE values reflect not only the
efficiency of H atom donation by R−H to cumyloxyl (k ,
a
Scheme 1) but also the efficiency of H atom atom transfer to
adduct radical intermediates (k , Scheme 2). Additionally, steric
tr
effects on macroradical addition to monomer (k , Scheme 2)
add
may influence the graft modification reactivity of the polymers
of interest. Since steric effects act on H atom atom transfer and
monomer addition in a parallel manner, a high AE measure-
ment may be indicative of an efficient graft propagation
sequence. Given sufficient monomer, these benefits will
improve kinetic chain lengths and VTMS conversions. As
such, AE may serve as a proxy measure of polymer reactivity
that is superior to oxidizability. In the present case, both AE
and VTMS conversion follow the same order: PEO > PE > PP
>
PIB. As noted above, PP is oxidized much faster than PE,
making oxidizability unsuitable as an indicator of the
amenability of a polymer toward graft modification chemistry.
Regioselectivity of H Atom Atom Transfer. The goal of
our model compound study was to improve our understanding
of the disparity in H atom atom donation reactivity of PE, PP,
and PIB at industrial reaction temperatures. Pentane, 2,4-
dimethylpentane (DMP), and 2,2,4,4-tetramethylpentane
(
TMP) contain the key structural elements of PE, PP, and
PIB, respectively. AE and VTMS conversion data were in good
agreement for the DMP-PP and TMP-PIB pairings, while
tetradecane reflected the reactivity of PE better than pentane
the three hydrocarbons, pentane afforded the highest
abstraction efficiency of 0.45, meaning that 45% of cumyloxyl
(
Table 1). van Drumpt and Oosterwijk report similar findings
for tert-butoxyl acting upon a homologous series of alkanes,
with AE increasing with chain length owing to higher
radicals engaged in H atom atom abstraction from the alkane as
•
opposed to fragmenting to CH + acetophenone. The product
3
48
proportions of methylene versus methyl groups. Nevertheless,
pentane was chosen for further study, since it generates a
limited number of positional isomers whose yields can, in turn,
be compared directly with those produced by DMP and TMP.
First used in the polymer context to examine the initiation of
methyl acrylate polymerization by tert-butoxyl, the trapping of
carbon-centered radicals with nitroxyls such as 1,1,3,3-
tetramethylisoindolin-2-oxyl (TMIO) has been adapted for
the purposes of H atom atom abstraction from alkanes,
generating important insight into H atom donation by
of AE and the isomer yield gives the percent of cumyloxyl
radicals that abstract from a given position. Whereas just 6% of
cumyloxyl radicals reacted at a methyl group, 25% and 14%
abstracted from the methylene groups at the 2-position and 3-
position, respectively. These positional abstraction efficiencies
can be normalized with respect to the number of functional
groups to account for statistical differences. Abstraction from a
methyl group can be characterized with an AE*yield/group
ratio of 3, whereas abstraction from methylene functionality is
defined by a ratio in the range of 13−14.
For DMP, an AE of 0.34 was determined, with the greatest
positional abstraction efficiency observed for the tertiary
position. 22% of cumyloxyl radicals reacted at this site,
producing a normalized AE*yield/group value of 11. It is
known that, on a per H atom basis, the tertiary position of
62
3
6,63
substrates such as DMP and 3-methylpentane.
Provided
that the resulting alkoxyamines are sufficiently stable with
respect to disproportionation, their relative yields reflect the
radical population established by initial H atom transfer events
(
Scheme 1). The primary and secondary alkoxyamines isolated
5
49
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
3
6,34
DMP is most reactive.
We note, however, that on a per
and entropic effects differs significantly between alkoxyl and
alkylperoxyl radicals, making oxidizability a poor predictor of a
polymer’s response to peroxide activation as well as vinyl-
trialkoxysilane grafting. We suggest that abstraction efficiency is
a more reliable indicator of the amenability of a polymer to
graft modification.
functional group basis, the tertiary position is no more reactive
than the secondary position of pentane. Moreover, the
normalized abstraction efficiency from the 3-position of DMP
was 3, compared to 14 for pentane. This 4-fold loss in reactivity
of the methylene group accounts for most of the difference in
the AE values of DMP and pentane. We suggest that this is the
underlying cause of the disparity in PE and PP reactivity toward
cumyloxyl.
TMP presents an even more extreme example of the
inhibitory effects imposed by methyl substituents. An overall
AE of 0.18 is provided entirely by primary H atom abstraction,
whose yield per group of 3 is consistent with that observed for
the methyl substituents of pentane and DMP. Reaction at the
The H atom atom transfer reactions of present interest are,
in principle, amenable to gas-phase quantum chemical
calculations, and several attempts to determine the structures
and free energies of relevant transition states have been
reported. An early attempt to calculate the overall thermody-
•
namic and thermokinetic parameters for t-BuO reactivity
confirmed that H atom abstraction from unhindered C−H
69
bonds is under enthalpic control. However, calculated
activation enthalpies were significantly different from accepted
experimental values. Similar difficulties are described by Bertin
et al., who used density functional theory (DFT) to characterize
3
-position is suppressed to such an extent that a secondary
alkoxyamine is produced only in trace quantities. This result
confirms speculation in an early report by Loan, in which the
primary sites within poly(isobutylene-co-isoprene), or IIR, were
suggested to be more reactive than secondary positions. The
regioselectivity of H atom transfer from the allylic positions
within IIR is unknown but is the subject of ongoing studies.
56
•
the reaction of t-BuO with a polypropylene model
5
3
compound. This level of theory provided qualitative
information regarding transition state energies, predicting that
H atom abstraction from tertiary positions faces entropic
barriers that are greater than enthalpic contributions to free
energies. That is, activation of the secondary positions within
DISCUSSION
■
Abstraction efficiency results, combined with regioselectivity
data, have provided insight into H atom atom abstractions from
a range of polymers under conditions that are relevant to
industrial practice. On the basis of low-temperature model
compound studies, Camera et al. concluded that H atom atom
abstraction from unhindered C−H bonds mirrors their BDE’s,
allylic > tertiary > secondary > methyl, whereas the reactivity of
hindered C−H bonds follows the order tertiary > methyl >
secondary. Our results are consistent with this view. Where
steric effects were equivalent, abstraction efficiencies for
CH − groups reflected differences in reaction enthalpy
PBD > PEO > PE). However, entropic effects controlled
•
PP by t-BuO is made difficult by the steric demands of the
adjactent tertiary positions.
Interestingly, subsequent DFT calculations on PP degrada-
tion confirmed that the reactivity of alkylperoxyl radicals with
respect to this hydrocarbon differs from that of alkoxy radicals,
in that enthalpic barriers are greater than entropic contributions
to transition state energies, leading to a kinetic preference for
reaction that follows the thermodynamics of the reaction; that
is, tertiary C−H abstraction is favored over secondary C−H
abstraction, which is favored over abstraction from the primary
3
4
−
(
2
6
8
the reactivity of −CH − groups bearing adjacent tertiary (PP)
position. This is consistent with our assertion that
oxidizability is not a reliable indicator of polypropylene’s
reactivity toward the high-energy alkoxyl and alkyl radical
intermediates that underlie typical nonoxidative cross-linking,
degradation and chemical modification processes.
2
and quaternary carbons (PIB), with steric inhibition of these H
atom donors accounting for differences in overall abstraction
efficiencies (PE > PP > PIB).
As noted above, H atom atom abstraction by cumyloxyl from
hydrocarbons is exergonic, with early transition states whose
structures and energetics are relatively insensitive to the overall
thermodynamics of possible reaction pathways. This results in a
weak correlation between BDE’s and cumyloxyl abstraction
efficiencies that can only be observed where reagents present
similar entropic barriers. In contrast, H atom abstraction by
alkylperoxyl radicals is endergonic, which is attributable to the
lower alkyl hydroperoxide OH bond strength (tert-butyl
CONCLUSIONS
■
The reactivity of cumyloxyl radicals at temperatures relevant to
industrial practice is entropy-controlled where substituents
present a steric barrier to H atom atom transfer. Therefore, PP
is a less reactive H atom donor than is PE, despite the low BDE
of its tertiary positions, owing to steric hindrance of abstraction
from the secondary C−H bonds. The reactivity if PIB is limited
almost exclusively to methyl group activation, with secondary
macroradical generation inhibited almost entirely by adjacent
52
hydroperoxide 352.3 ± 8.8 kJ/mol) compared to that of
the corresponding alcohol (444.9 ± 2.8 kJ/mol), with
52
65
correspondingly later transition states. Therefore, differences
in the overall reaction enthalpy, as measured through BDE’s,
are expressed more strongly in the transition state free energies,
the result being greater regioselectivity for tertiary over
secondary and primary C−H bonds. This is not to say that
entropic effects do not contribute to oxidizability, as steric
hindrance is often invoked to account for the relative stability of
polymers such as polystyrene and poly(vinylcylcohexane)
•
quaternary carbons. Differences in the reactivity of ROO and
•
RO in H atom transfer are significant enough to render
oxidizability a poor measure of the amenability of a polyolefin
toward radical-mediated chemical modification.
66
AUTHOR INFORMATION
■
67
toward oxidation. However, the presence of an extra oxygen
atom in alkylperoxyl radicals increases the distance between the
abstraction site and the steric bulk on the radical, reducing
Corresponding Author
*
E-mail parent@chee.queensu.ca; Ph (613)533-6266 (J.S.P.).
68
Notes
entropic effects compared to alkoxyl radicals. Given the
foregoing, it would appear that the balance between enthalpic
The authors declare no competing financial interest.
5
50
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551

Macromolecules
Article
(
38) Griffiths, P. G.; Moad, G.; Rizzardo, E.; Solomon, D. H. Aust. J.
ACKNOWLEDGMENTS
■
Chem. 1983, 36, 397.
39) Ogle, C. A.; Martin, S. W.; Dziobak, M. P.; Urban, M. W.;
Financial support from E.I. DuPont Canada and the Natural
Sciences and Engineering Research Council (NSERC) is
gratefully acknowledged.
(
Mendenhall, G. D. J. Org. Chem. 1983, 48, 3728.
(40) Griffiths, P. G.; Rizzardo, E.; Solomon, D. H. J. Macromol. Sci.,
Part A: Chem. 1982, 17, 45.
(
41) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317.
REFERENCES
■
(42) Rowe, P. D.; Thomas, D. K. J. Appl. Polym. Sci. 1963, 7, 461.
(43) Fittschen, C.; Hippler, H.; Viskolcz, B. Phys. Chem. Chem. Phys.
2000, 2, 1677.
(
(
(
(
1) Chodak, I. Prog. Polym. Sci. 1995, 20, 1165.
2) Lazar, M.; Rado, R.; Rychly, J. Adv. Polym. Sci. 1990, 95, 151.
3) Hyslop, D. K.; Parent, J. S. Macromolecules 2012, 45, 8147.
4) Tzoganakis, C.; Vlachopoulos, J.; Hamielec, A. E. Polym. Eng. Sci.
989, 29, 390.
5) Ryu, S. H.; Gogos, C. G.; Xanthos, M. Adv. Polym. Technol. 1992,
1, 121.
́
(44) Berces, T.; Trotman-Dickenson, A. F. J. Chem. Soc. 1961, 384.
(45) Minisci, F.; Galli, R. Tetrahedron Lett. 1962, 12, 533.
(46) Ingold, K. U. In Free Radicals; Kochi, J. K., Ed.; Wiley: New
York, 1973; Vol. 1, p 92.
1
(
1
(
(47) Zytowski, T.; Fischer, H. J. Am. Chem. Soc. 1997, 119, 12869.
(48) Van Drumpt, J. D.; Oosterwijk, H. H. J. Polym. Sci., Polym. Chem.
Ed. 1976, 14, 1495.
6) Moad, G. Prog. Polym. Sci. 1999, 24, 81.
7) Munteanu, D. In Reactive Modifiers for Polymers; Al-Malaika, S.,
(
Ed.; Blackie Academic and Professional: London, 1997; p 196.
(49) Zhou, W.; Zhu, S. Ind. Eng. Chem. Res. 1997, 36, 1130.
(50) Parent, J. S.; Parodi, R.; Wu, W. Polym. Eng. Sci. 2006, 46, 1754.
(51) McMillan, D. E.; Golden, P. M. Annu. Rev. Phys. Chem. 1982, 33,
493.
(8) Liu, N. C.; Baker, W. E. Adv. Polym. Technol. 1992, 11, 249.
(9) Passaglia, E.; Coiai, S.; Augier, S. Prog. Polym. Sci. 2009, 34, 911.
(10) El Mabrouk, K.; Parent, J. S.; Chaudhary, B. I.; Cong, R. Polymer
2
(
009, 50, 5390.
(52) Luo, Y. R. In Comprehensive Handbook of Chemical Bond
Energies; CRC Press: Boca Raton, FL, 1997.
11) Dluzneski, P. R. Rubber Chem. Technol. 2000, 74, 451.
(
12) Tzoganakis, C.; Tang, Y.; Vlachopoulos, J.; Hamielec, A. E.
(53) Bertin, D.; Grimaldi, S.; Leblanc, M.; Marque, S. R. A.; Siri, D.;
Tordo, P. J. Mol. Struct: THEOCHEM 2007, 811, 255.
(54) Dokolas, P.; Solomon, D. H. Polymer 2000, 41, 3523.
(55) Thomas, D. K. Trans. Faraday Soc. 1961, 57, 511.
(56) Loan, L. D. J. Polym. Sci., Part A: Polym. Chem. 1964, 2, 2127.
(57) Mixer, R. Y.; Bailey, D. L. J. Polym. Sci. 1955, 18, 573.
(58) Seyferth, D.; Soon, D. Y.; Tasi, M. Organometallics 1995, 14,
2225.
(59) Forsyth, J. C.; Baker, W. E.; Russell, K. E.; Whitney, R. A. J.
Polym. Sci., Part A: Polym. Chem. 1997, 35, 3517.
(60) Spencer, M.; Parent, J. S.; Whitney, R. A. Polymer 2003, 44,
2015.
Polym.-Plast. Technol. Eng. 1989, 28, 319.
(
(
(
(
13) Russell, K. E. Prog. Polym. Sci. 2002, 27, 1007.
14) Parent, J. S.; Parodi, R.; Wu, W. Polym. Eng. Sci. 2006, 46, 1754.
15) Lazar, M.; Rychly, J. Adv. Polym. Sci. 1992, 102, 190.
16) Lloyd, W. G. In Methods in Free-Radical Chemistry; Huyser, E. S.,
Ed.; Marcel Dekker: New York, 1973; Vol. 4, p 1.
17) Norling, P. M.; Lee, T. C. P.; Tobolsky, A. V. Rubber Chem.
Technol. 1965, 38, 1198.
18) Grassie, N.; Scott, G. In Polymer Degradation and Stabilization;
Cambridge University Press: New York, 1988; p 92.
19) Van Sickle, D. E. J. Polym. Sci., Part A: Polym. Chem. 1972, 10,
75.
(
(
(
2
(
61) Parent, J. S.; Tripp, M.; Dupont, J. Polym. Eng. Sci. 2003, 43,
34.
62) Cuthbertson, M. J.; Rizzardo, E.; Solomon, D. H. J. Macromol.
Sci., Chem. 1982, 17, 45−50.
63) Dokolos, P.; Looney, M. G.; Musgrave, S.; Poon, S.; Solomon,
D. H. Polymer 2000, 41, 3137.
64) Scott, M. E.; Parent, J. S.; Dupont, J.; Whitney, R. A. Ind. Eng.
Chem. Res. 2003, 42, 3662.
65) Howard, J. A. In Free Radicals; Kochi, J. K., Ed.; Wiley: New
York, 1973; Vol. 2, p 18.
2
(
(20) Beachell, H. C.; Beck, D. L. J. Polym. Sci., Part A 1965, 3, 457.
(21) Iring, M.; Tudos, F. Prog. Polym. Sci. 1990, 15, 217−262.
(22) Antunes, F.; Barclay, L .R. C.; Ingold, K. U.; King, M.; Norris, J.
(
Q.; Scaiano, J. C.; Xi, F. Free Radical Biol. Med. 1999, 26, 117.
(
(
23) Ross, L.; Barclay, C.; Vinqvist, M. R. Org. Lett. 2000, 2, 2841.
24) Finn, M.; Friedline, R.; Suleman, N. K.; Wohl, C. J.; Tanko, J.
(
M. J. Am. Chem. Soc. 2004, 126, 7578.
25) Avila, D. V.; Brown, C. E.; Ingold, K. U.; Lusztyk, J. J. Am. Chem.
Soc. 1993, 115, 466.
(
(
(
(
66) Middleton, B. S.; Ingold, K. U. Can. J. Chem. 1967, 45, 191.
67) Hansen, R. H. In Thermal Stability of Polymers; Conley, R. T.,
(
(
26) Sakurai, H.; Hosomi, A. J. Am. Chem. Soc. 1967, 89, 458.
27) Lal, L.; McGrath, J. E.; Board, R. D. J. Polym. Sci., Part A 1968, 6,
Ed.; Marcel Dekker: Ann Arbor, MI, 1970; Vol. 1, p 157.
68) Bertin, D.; Leblanc, M.; Marque, S. R. A.; Siri, D. Polym. Degrad.
Stab. 2010, 95, 782.
69) Opeida, I. A.; Dmitruk, A. F.; Zarechnaya, O. M. Russ. Chem.
Bull. Int. Ed. 2001, 50, 241.
8
(
4
(
21.
28) Chodak, I.; Bakos, D. Coll. Chechoslov. Chem. Commun. 1978,
3, 2574.
29) (a) Niki, E.; Kamiya, Y.; Ohta, N. Bull. Chem. Soc. Jpn. 1968, 41,
466. (b) Smith, J. R. L.; Nagatomi, E.; Stead, A.; Waddington, D. J.;
(
(
1
Beviere, S. D. J. Chem. Soc., Perkin Trans. 2 2000, 1193. (c) Baignee,
A.; Howard, J. A.; Sciano, J. C.; Stewart, L. C. J. Am. Chem. Soc. 1983,
1
(
1
(
05, 6120.
30) (a) Niki, E.; Kamiya, Y. J. Chem. Soc., Perkin Trans. 2 1975,
221. (b) Niki, E.; Kamiya, Y. J. Org. Chem. 1973, 38, 1403.
31) Forsyth, J. C.; Baker, W. E.; Russell, K. E.; Whitney, R. A. J.
Polym. Sci., Part A: Polym. Chem. 1997, 35, 3517.
32) Heinen, W.; Rosenmoller, C. H.; Wenzel, C. B.; de Groot, H. J.
M.; Lugtenburg, J.; van Duin, M. Macromolecules 1996, 29, 1151.
33) Niki, E.; Ohto, N.; Kanauchi, T.; Kamiya, Y. Eur. Polym. J. 1980,
6, 559.
34) Camara, S.; Gilbert, B. C.; Meier, R. J.; van Duin, M.;
Whitwood, A. C. Org. Biomol. Chem. 2003, 1, 1181.
(
(
1
(
(
35) Rizzardo, E.; Solomon, D. H. Polym. Bull. 1979, 1, 529.
(
36) Dokolas, P.; Loffler, S. M.; Solomon, D. Aust. J. Chem. 1988, 51,
1113.
(
37) Mendenhall, G. Tetrahedron Lett. 1983, 24, 451.
5
51
dx.doi.org/10.1021/ma402177v | Macromolecules 2014, 47, 544−551