Isobutylene-rich macromonomers: Dynamics and yields of peroxide-initiated crosslinking

Article abstract of DOI:10.1002/pola.27462

New isobutylene-rich elastomers bearing multiple pendant styrenic, acrylic, maleimidic, vinylic, and allylic functional groups have been prepared and examined in the context of peroxide-initiated crosslinking. Halide displacement from brominated poly(isobutylene-co-isoprene) (BIIR) by the requisite carboxylate nucleophiles in homogeneous toluene solutions provide the desired esters in quantitative yield without complications from dehydrohalogenation or premature crosslinking. Heating the resulting macromonomers with dicumyl peroxide to 160 °C under solvent-free conditions gives thermoset derivatives, with reaction rates and yields depending markedly on functional group structure. In general, high cure extents can only be achieved using highly reactive pendant functional groups, owing to the competitive balance between crosslinking through C=C oligomerization, and degradation through β-scission of backbone macroradical intermediates. Independent control of crosslinking rates and cure extents is gained through the use of nitroxyl radical traps bearing acrylate functionality.

Full text of DOI:10.1002/pola.27462

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Isobutylene-Rich Macromonomers: Dynamics and Yields of
Peroxide-Initiated Crosslinking
Jackson M. Dakin,1 Karthik Vikram Siva Shanmugam,1 Christopher Twigg,1
Ralph A. Whitney,2 J. Scott Parent1
1Department of Chemical Engineering, Queen’s University, Kingston, Ontario K7L 3N6, Canada
2Department of Chemistry, Queen’s University, Kingston, Ontario K7L 3N6, Canada
Correspondence to: J.S. Parent (E-mail: parent@queensu.ca)
Received 16 September 2014; accepted 19 October 2014; published online 00 Month 2014
DOI: 10.1002/pola.27462
ABSTRACT: New isobutylene-rich elastomers bearing multiple
pendant styrenic, acrylic, maleimidic, vinylic, and allylic func-
tional groups have been prepared and examined in the context
of peroxide-initiated crosslinking. Halide displacement from
brominated poly(isobutylene-co-isoprene) (BIIR) by the requi-
site carboxylate nucleophiles in homogeneous toluene solu-
tions provide the desired esters in quantitative yield without
complications from dehydrohalogenation or premature cross-
linking. Heating the resulting macromonomers with dicumyl
peroxide to 160 ꢀC under solvent-free conditions gives thermo-
set derivatives, with reaction rates and yields depending mark-
edly on functional group structure. In general, high cure
extents can only be achieved using highly reactive pendant
functional groups, owing to the competitive balance between
crosslinking through C@C oligomerization, and degradation
through b-scission of backbone macroradical intermediates.
Independent control of crosslinking rates and cure extents is
gained through the use of nitroxyl radical traps bearing acry-
C
V
late functionality. 2014 Wiley Periodicals, Inc. J. Polym. Sci.,
Part A: Polym. Chem. 2014, 00, 000–000
KEYWORDS: butyl rubber; elastomers; polymer modification;
radical crosslinking
underlies peroxide-cures of saturated polymers.7 It should
be noted that macromonomers are often defined as materials
bearing polymerizable end-groups,8–10 whereas in the pres-
ent context, BIIR-derived macromonomers contain many pol-
ymerizable groups distributed randomly along the polymer
chain.
INTRODUCTION Poly(isobutylene-co-isoprene) (IIR) is a ran-
dom copolymer that is prepared by cationic polymerization
of isobutylene and 1–2 mol % isoprene. This elastomeric
material owes its superior oxidative stability and gas imper-
meability to its isobutylene-rich polymer backbone,1 whereas
the residual unsaturation provided by isoprene isomers sup-
ports the sulfur vulcanization reactivity needed to transform
IIR into thermoset articles.2 Cure reactivity is significantly
improved by halogenating isoprene-derived unsaturation to
yield brominated poly(isobutylene-co-isoprene) (BIIR),3
whose allylic halide functionality is highly reactive to sulfur
nucleophiles.4 Unfortunately, sulfur-cure formulations require
accelerators and co-curing agents, resulting in a vulcanized
product containing leachable and extractable residues.
An early example of an isobutylene-rich macromonomer was
described by Oxley and Wilson, who polymerized isobuty-
lene, isoprene, and divinylbenzene to give an elastomer that
cured by radical oligomerization of residual styrenic func-
tionality.11 However, the activation of divinylbenzene during
the cationic polymerization process yielded a gel content of
85% that compromised the material’s processing characteris-
tics. Wang et al. prepared gel-free analogs by postpolymeri-
zation modification of brominated poly(isobutylene-co-para-
methylstyrene), yielding acrylate derivatives that cured to
high extent when activated by peroxide thermolysis, UV-
initiation, and electron beam bombardment.12 More recently,
BIIR-derived macromonomers have been prepared by the
Suzuki-Miyaura coupling reaction using 4-vinylphenylboronic
acid,13 and through bromide displacement by acrylate and
vinyl benzoate nucleophiles.7
Although peroxide-based cure formulations are generally
much cleaner, IIR incurs radical degradation when activated
by organic peroxides.5,6 This limitation can be overcome by
introducing small amounts (between 0.05 and 0.20 mmol/g)
of polymerizable functionality to the material to generate
macromonomer derivatives that crosslink by radical homo-
polymerization of pendant groups, as opposed to the hydro-
gen abstraction 1 macro-radical combination mechanism that
C
V
2014 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
1

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
tes > methacrylates > styrenics> itaconates.14 However, reac-
tion conditions used in elastomer curing are substantially
ꢀ
different, with temperatures in the range of 160 C, and poly-
merizable functionalities rarely exceeding concentrations of
0.25 mmoles/g polymer. Moreover, several of the functional
groups of interest have received little attention, such as vinyl
and allyl imidazolium bromide ionic liquids.
This report describes the synthesis of several new IIR-based
macromonomers (Sch. 1), and characterizes their reactivity
toward peroxide-initiated crosslinking under conditions that
are relevant to industrial practice. Reaction dynamics and
yields are measured by dynamic shear rheometry, and
results are discussed in the context of underlying polymer
crosslinking and scission mechanisms. The study concludes
with a demonstration of new chemistry for delaying the
onset of macromonomer crosslinking without compromising
ultimate cure yields.
EXPERIMENTAL
Materials
Acrylic acid (99%), vinyl benzoic acid (VBA, 97%), metha-
crylic acid (MAA, 99%), 10-undecenoic acid (UA, 98%), maleic
anhydride (99%), 4-aminobenzoic acid (99%), sodium acetate
(anhydrous, 99%), acetic anhydride (99%), itaconic anhydride
(95%), dodecanol (95%), tetrabutylammonium bromide
(TBAB, 98%), tetrabutylammonium hydroxide (Bu4NOH, 1M
solution in methanol, 98%), dicumyl peroxide (DCP, 98%),
triethylamine (Et3N 99%), N,N’-m-phenylene maleimide
(98%), 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO, sublimed
99%), 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl (TEM-
POH, 97%), and 1-vinylimidazole (VIm, 991%) were used as
received from Sigma-Aldrich. 1-Allylimidazole (AlIm, 99%)
was used as received from Alfa Aesar. Butyl rubber (IIR,
2.8 mol % isoprene, RB-301) and BIIR (Bromobutyl 2013,
allylic bromide content ꢁ0.15 mmol/g, Mn ꢂ 400 K) were
used as received by LANXESS Inc. (Sarnia, ON).
Synthesis of IIR-g-Methacrylate (IIR-MAA)
Methacrylic acid (0.322 g, 3.7 mmol) was treated with a 1M
solution of Bu4NOH in methanol (3.7 mL, 3.7 mmol Bu4NOH)
to yield the desired Bu4Ncarboxylate salt, which was isolated
by removing methanol under vacuum. BIIR (11 g) and
Bu4NBr (0.53 g, 1.65 mmol) were dissolved in toluene
SCHEME 1 Reactive functional groups within BIIR-derived
macromonomers.
ꢀ
(100 g) and heated to 85 C for 180 min. The Bu4Ncarboxy-
late salt (1.2 g, 3.7 mmol) was added before heating the
reaction mixture to 85 ꢀC for 60 min. The esterification
product was isolated by precipitation from excess acetone,
purified by dissolution/precipitation using THF /acetone,
and dried under vacuum, yielding IIR-g-methacrylic acid
(IIR-MAA). 1H-NMR (CDCl3): d 6.03 (s,CH2@C(CH3)–COO-,
The current work is concerned with the relationship between
the structure of a macromonomer’s functionality and the mate-
rial’s crosslinking dynamics and yields. Scheme 1 illustrates
the macromonomers of interest, which range from highly reac-
tive acrylates and styrenics,14 to relatively unreactive maleate
diesters15–17 and unactivated allylic monomers.18,19 Some of
these functional groups have been studied extensively in the
context of low temperature homopolymerization, with reliable
estimates of the propagation rate constants providing insight
into the potential reactivity of a given macromonomer system.
For example, rate constants measured at 60 ꢀC suggest that
macromonomer reactivity may decline in the order acryla-
1H),
d
5.47 (s,CH2@C(CH3)–COO-,1H),
d
3.36 (s,
(CH2@C(CH3)–COO-, 3H),d 4.51 (E-ester, @CH–CH2–OCO-,
2H, s), d 4.56 (Z-ester, @CH–CH2–OCO–, 2H, s).
Synthesis of IIR-g-Maleimido Benzoate (IIR-MBA)
Maleic anhydride (3 g, 0.03 mol) and 4-aminobenzoic acid
(4.2 g, 0.03 mol) were dissolved in acetone (10 g) and the
2
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
mixture was stirred for 12 h at room temperature. The insol-
uble amidic acid precipitate was filtered off and dried in vac-
uum. The amidic acid (6 g, 0.025 mol) was treated with
acetic anhydride (8 g, 0.078 mol) and sodium acetate (1 g,
0.012 mol) at 80 ꢀC for 15 min. Maleimidobenzoic acid was
isolated by precipitating through successive precipitation-
dissolution in water/THF and dried in vacuum. 1H-NMR
(DMSO-d6): d 7.21 (s,-CH@CH-, 2H), d 7.50 (d, Ar–H, 2H), d
fication product was isolated by precipitation from excess
acetone, purified by dissolution/precipitation using THF/ace-
tone, and dried under vacuum, yielding IIR-g-undecenoate
1
(IIR-UA). H-NMR (CDCl3): d 5.68 (m,–CH2@CH–CH2–, 1H), d
4.85 (d, CH2@CH-, 1H), d 4.79 (d, CH2@CH–, 1H), d 4.51 (E-
ester, @CH–CH2–OCO-, 2H, s), d 4.56 (Z-ester, @CH–CH2–
OCO–, 2H, s).
Synthesis of IIR-g-Dodecyl Maleate (IIR-DDMA)
8.02 (d, Ar–H, 2H). HR MS calculated for
C11H7NO4:
A total of 1-Dodecanol (1.5 g, 8.04 mmol) and maleic anhy-
dride (0.98 g, 10.05 mmol) were dissolved in toluene (10 g)
217.0371; found for C11H7NO4: 217.0375.
ꢀ
BIIR (10 g), TBAB (0.48 g, 1.5 mmol), maleimidobenzoic acid
(0.651 g, 3.0 mmol) and triethylamine (0.3g, 3.0 mmol) were
dissolved in THF (100 g) and refluxed in a 250-mL vessel
for 600 min. The esterification product was isolated by pre-
cipitation from excess acetone, purified by dissolution/pre-
cipitation using THF/acetone, and dried under vacuum,
yielding IIR-g-maleimido-benzoate (IIR-MBA).
and heated to 80 C for 4 h. Residual starting materials and
solvent were removed by Kugelrohr distillation (T 5 80 ꢀC,
P 5 0.6 mmHg). The resulting acid-ester was isolated. 1H-
NMR (CDCl3): d 6.46 (d, HOOC–CH@CH–COO–, 1H), d 6.36
(d, HOOC–CH@CH–COO–, 1H), d 4.27 (t, @CH–COO–CH2–,
2H), d 1.71 (m, –COO–CH2–CH2–, 2H), d 1.36 (m, –COO–
(CH2)10–CH2–CH3, 2H), d 1.25 (m, –CH2–(CH2)9–CH2–CH3,
18H), d 0.87 (t, CH2–CH3, 3H). HR MS calculated for
Synthesis of IIR-g-Dodecyl Itaconate (IIR-DDI)
C16H29O4: 285.2078; found for C16H29O4: 285.2065.
Monododecyl itaconate was prepared as previously
described20.1-Dodecanol (8.0 mmol, 1.5 g) and itaconic anhy-
dride (24 mmꢀol, 2.7 g), were dissolved in toluene (10 g) and
heated to 80 C for 4 h. Residual starting materials and sol-
vent were removed by Kugelrohr distillation (T 5 80 ꢀC,
P 5 0.6 mmHg). The resulting acid-ester was isolated and
dried. 1H-NMR (DMSO-d6): d 6.14 (d, HOOC–C(@CH2)–CH2–
COO–, 1H), d 5.74 (d, HOOC–C(@CH2)–CH2–COO–, 1H), d
3.28 (s, HOOC–C(@CH2)–CH2–COO–, 1H) d 3.98 (t, –CH2–
COO–CH2–, 2H), d 1.52 (m, –COO–CH2–CH2–, 2H), d 1.36 (m,
-COO–(CH2)10–CH2–CH3, 2H), d 1.25 (m, –CH2–(CH2)9– CH2–
CH3, 18H), d 0.87 (t, CH2–CH3, 3H). HR MS calculated for
Monododecyl maleate (0.94 g, 3.3 mmol) was treated with a
1M solution of Bu4NOH in methanol (3.3 mL, 3.3 mmol
Bu4NOH) to yield the desired Bu4Ncarboxylate salt, which
was isolated by removing methanol under vacuum. BIIR
(11 g) and Bu4NBr (0.53 g, 1.65ꢀmmol) were dissolved in
toluene (100 g) and heated to 85 C for 180 min. The Bu4N-
carboxylate salt (1.7 g, 3.3 mmol) was added before heating
the reaction mixture to 85 ꢀC for 60 min. The esterification
product was isolated by precipitation from excess acetone,
purified by dissolution/precipitation using THF/acetone, and
dried under vacuum, yielding IIR-g-dodecyl maleate (IIR-
1
C17H31O4: 299.2225; found for C17H31O4: 299.2222.
DDMA). H-NMR (CDCl3): d 6.19 (s, OOC–CH@CH–COO–, 2H),
d 4.11 (t, –CH2–CH2–COO–CH2–, 2H), d 4.58 (E-ester, @CH–
CH2–OCO-, 2H, s), d 4.66 (Z-ester, @CH–CH2–OCO–, 2H, s).
Monododecyl itaconate (0.98 g, 3.3 mmol) was treated with
a 1M solution of Bu4NOH in methanol (3.3 mL, 3.3 mmol
Bu4NOH) to yield the desired Bu4Ncarboxylate salt, which
was isolated by removing methanol under vacuum. BIIR
(11 g) and Bu4NBr (0.53 g, 1.65ꢀmmol) were dissolved in
toluene (100 g) and heated to 85 C for 180 min. The Bu4N-
carboxylate salt (1.78 g, 3.3 mmol) was added before heating
the reaction mixture to 85 ꢀC for 60 min. The esterification
product was isolated by precipitation from excess acetone,
purified by dissolution/precipitation using THF/acetone, and
dried under vacuum, yielding IIR-g-dodecyl itaconate (IIR-
DDI). 1H-NMR (CDCl3): d 6.24 (s, CH2@C(CH2)–COO–, 1H), d
5.62 (s, CH2@C(CH2)–COO–, 1H), d 3.36 (s, CH2@C(CH2)–
COO–, 2H), d 4.54 (E-ester, @CH–CH2–OCO–, 2H, s), d 4.60
(Z-ester, @CH–CH2–OCO–, 2H, s).
Synthesis of IIR-g-Vinyl benzoate (IIR-VBA)
A total of 4-Vinylbenzoic acid (0.44 g, 3.0 mmol) was
treated with a 1M solution of Bu4NOH in methanol (3.0 mL,
3.0 mmol Bu4NOH) to yield the desired Bu4Ncarboxylate
salt, which was isolated by removing methanol under vac-
uum. BIIR (10 g) and Bu4NBr (0.77 g, 2.4 mmol) were dis-
solved in toluene (100 g) and heated to 85 ꢀC for 180 min.
The Bu4Ncarboxylate salt (1.16 g, 3.0 mmol) was added
before heating the reaction mixture to 85 ꢀC for 60 min.
The esterification product was isolated by precipitation
from excess acetone, purified by dissolution/precipitation
using THF /acetone, and dried under vacuum, yielding
IIR-g-vinyl benzoate (IIR-VBA). 1H-NMR (CDCl3): d 6.73
(dd,-CH2@CH–Ar, 1H), d 5.82 (dd, CH2@CH-, 1H), d 5.48
(dd, CH2@CH–, 1H), d 4.78 (E-ester, @CH–CH2–OCO–, 2H, s),
d 4.81 (Z-ester, @CH–CH2–OCO–, 2H, s).
Synthesis of IIR-g-Undecenoate (IIR-UA)
A total of 10-Undecenoic acid (0.88 g, 3.7 mmol) was treated
with a 1M solution of Bu4NOH in methanol (4.8 mL, 4.8
mmol Bu4NOH) to yield the desired Bu4Ncarboxylate salt,
which was isolated by removing methanol under vacuum.
BIIR (16 g) and Bu4NBr (0.77 g, 2.4 mmol) were dissolved
in toluene (100 g) and heated to 85 ꢀC for 180 min. The
Bu4Ncarboxylate salt (2.04 g, 4.8 mmol) was added before
Synthesis of IIR-g-Acrylate (IIR-AA)
Acrylic acid (0.21 g, 3.0 mmol) was treated with a 1M solu-
tion of Bu4NOH in methanol (3.0 mL, 3.0 mmol Bu4NOH) to
yield the desired Bu4Ncarboxylate salt, which was isolated
by removing methanol under vacuum. BIIR (10 g) and
Bu4NBr (0.48 g, 1.5 mmol) were dissolved in toluene
ꢀ
heating the reaction mixture to 85 C for 60 min. The esteri-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
3

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
ꢀ
(100 g) and heated to 85 C for 180 min. The Bu4Ncarboxy-
late salt (0.94 g, 3.0 mmol) was added before heating the
reaction mixture to 85 ꢀC for 60 min. The esterification
product was isolated by precipitation from excess acetone,
purified by dissolution/precipitation using THF/acetone, and
dried under vacuum, yielding IIR-g-acrylate (IIR-AA).
1H-NMR (CDCl3): d 6.12 (dd, CH2@CH–COO–, 1H), d 6.40
(dd, CH2@CH–, 1H), d 5.82 (dd, CH2@CH-, 1H), d 4.66
(E-ester, @CH–COO–CH2–, 2H, s), d 4.60 (Z-ester, @CH–COO–
CH2–, 2H, s).
Synthesis of IIR-g-Vinylimidazolium Bromide IIR-VImBr
BIIR (20 g, 3.0 mmol allylic bromide) was dissolved in tolu-
ene (200 mL) and heated to 100 ꢀC. 1-Vinylimidazole
(1.69 g, 18.0 mmol, 6 eq.) was then added to the solution
and the mixture was allowed to react for 50 h. The
N-alkylation product was isolated by precipitation from
excess acetone, purified by dissolution/precipitation using
THF/acetone, and dried under vacuum, yielding IIR-g-
vinylimidazolium bromide (IIR-VImBr). 1H-NMR (CDCl3 1 5
wt % CD3OD): d 11.53 (s, -N1–CH–N-, 1H), d 7.61 (s, 1H,
-N1–CH@CH–N-), d 7.42 (dd,1H, -N–CH@CH2), d 5.80 (dd,
1H, N–CH@CH–Htrans), d 5.41 (dd, 1H, N–CH@CH–Hcis),
found for the Z-isomer: d 7.16 (s, 1H, Z, sbond]N1–CH@CH–
N–), d 5.69 (t, 1H, Z, CH2–C@CH–CH2) d4.91 (s, 2H, -C–CH2–
N1), found for the E-isomer: d 7.20 (s, 1H, E, -N1-CH@CH–
N–), d 5.68 (t, 1H, E, CH2–C@CH–CH2), d 4.84 (s, 2H, E, –C–
CH2–N1). 1H-NMR for conjugated dienes: d 5.98 (d, 1H,
exo-diene), d 5.92 (d, 1H, endo-diene). 1H-NMR for residual
isoprene mer: d 5.04 (t, 1H, CH3–C@CH–CH2-).
SCHEME 2 Radical reactions of polyisobutylene.
QqTOF mass spectrometer. Dynamic shear modulus measure-
ments were recorded using an Alpha Technologies, Advanced
Polymer Analyzer 2000. Pressed samples of elastomer were
coated with the required amount of a stock solution of DCP
in acetone, and allowed to dry before passing three times
through a 2-roll mill. Compounds containing TEMPO or
AOTEMPO were prepared similarly by coating the polymer
with solutions of DCP and nitroxyl before allowing to dry for
30 min in air before mill mixing. Care was made to ensure
that TEMPO losses to sublimation were minimized. The com-
pound was then cured at 160 ꢀC in the rheometer cavity at
3ꢀ oscillation arc and a frequency of 1 Hz.
Synthesis of IIR-g-Allylimidazolium Bromide IIR-AlImBr
BIIR (20 g, 3.0 mmol allylic bromide) was dissolved in tolu-
ꢀ
ene (200 mL) and heated to 100 C. 1-Allylimidazole (1.69 g,
18.0 mmol, 6 eq.) was then added to the solution and the
mixture was allowed to react for 24 h. The N-alkylation
product was isolated by precipitation from excess acetone,
purified by dissolution/precipitation using THF/acetone, and
dried under vacuum, yielding IIR-g-allylimidazolium bromide
(IIR-AlImBr). 1H-NMR (CDCl3 1 5 wt % CD3OD): d 9.79 (s, -
N1–CH–N–, 1H), d 7.21 (s, 1H, -N1–CH@CH–N–), d 6.00
(ddt,1H, –N–CH2–CH@CH2),
d
5.40 (dd, 1H, N–CH2–
CH@CH2), found for the Z-isomer: d 7.12 (s, 1H, Z, -N1-
CH@CH–N-), d 5.66 (t, 1H, Z, CH2–C@CH–CH2) d4.85 (s, 2H,
-C–CH2–N1), found for the E-isomer: d 7.14 (s, 1H, E, -N1–
CH@CH–N-), d 5.62 (t, 1H, E, CH2–C@CH–CH2), d 4.78 (s, 2H,
RESULTS AND DISCUSSION
Butyl Rubber Degradation
The challenge in developing peroxide-cure formulations for
isobutylene-rich elastomers is to overcome the susceptibility
of the polymer backbone toward radical degradation. Hydro-
gen atom abstraction from polyisobutylene by cumyloxy rad-
icals is regioselective for primary alkyl positions, due to
steric hindrance of secondary positions imposed by adjacent
quaternary carbons (Sch. 2).22 Termination of the resulting
primary macroradicals by combination can generate cross-
links. However, as noted by Loan, b-scission is a dominant
chain reaction, resulting in molecular weight losses when
polyisobutylene is heated with DCP to 153 ꢀC.6 Loan also
reports that butyl rubber (IIR) containing less than 2 mole
% isoprene is less susceptible to radical degradation than is
polyisobutylene, since H-atom abstraction from allylic
1
E, -C–CH2–N1). H-NMR for conjugated dienes: d 5.92 (d, 1H,
endo-diene). 1H-NMR for residual isoprene mer: d 5.03 (t,
1H, CH3–C@CH–CH2-).
Synthesis of AOTEMPO
The synthesis and purification of 4-acryloyloxy-2,2,6,6-
tetramethylpiperidine-N-oxyl (AOTEMPO) is described
elsewhere.21
Instrumentation and Analysis
1H-NMR spectra were acquired on a Bruker Avance-400
spectrometer; only downfield signals are reported. Mass
spectra were obtained on an Applied Biosystems QStar XL
4
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
transfer techniques have been developed to overcome this
issue,28 but the resulting reaction kinetics can be too slow
for the present application, since many of the macromono-
mers of interest are susceptible to cross linking during their
preparation. To accelerate the esterification process, we have
used tetrabutylammonium carboxylate salts as toluene-
soluble nucleophiles, and we isomerized the exomethylene
allylic bromide functionality in the starting material to more
kinetically more reactive E,Z-bromomethyl isomers by heat-
ꢀ
ing BIIR with Bu4NBr in toluene at 85 C. The higher electro-
philicity of the isomerized polymer,29 coupled with the use
of soluble carboxylate nucleophiles, allowed esterifications to
be completed in 1 h without complications associated with
BIIR dehydrohalogenation.30 The resulting macromonomers
contained 0.15 mmoles of reactive functionality per gram of
elastomer, as measured by 1H-NMR spectrum integration.
These functional groups are distributed randomly within
each polymer chain, and exist predominately in their Z-
isomeric form due to the unique selectivity of the pre-
isomerization process (Sch. 1).31 Since esterification is not
accompanied by changes in the polymer’s molecular
weight,26 given a number average molecular weight of
400,000 g/mol for the starting material, about 60 functional
groups per polymer chain were introduced, with an average
molecular weight between functional groups (Mc) on the
order of 6500 g/mol.
FIGURE 1 Dynamics of IIR degradation by DCP initiated chain
scission at increasing peroxide loadings, and in the presence
of unbound styrene.
positions yields allyl radical intermediates that prefer termi-
nation by combination to cleavage.
Figure 1 illustrates the degradation suffered by IIR when it
is heated to 160 ꢀC with varying amounts of DCP initiator.
These data are measurements of dynamic storage modulus
(G0) recorded at a fixed temperature, frequency, and shear
strain amplitude,23,24 and they demonstrate clear evidence of
chain scission. The timescale of the degradation process is
dictated by the first-order decomposition rate of the perox-
ide, whose 5.4 min half-life at 160 ꢀC25 leads to an overall
reaction time on the order of 40 min. In the context of
macro-monomer design, this polymer backbone fragmenta-
tion must be dominated by radical oligomerization of pend-
ant C@C functionality such that substantially all chains are
incorporated in a covalent network of adequate crosslink
density.
Instability in regards to auto-initiated crosslinking, even at
room temperature, has been reported for closely related sys-
tems.7,32,33 Wang et al. went so far as to graft antioxidant
functionality to their macromonomer system to improve its
storage stability.12 Our early work with IIR-MBA, IIR-VBA,
and IIR-AA was similarly complicated by crosslinking over a
period of days at ambient temperaturꢀe, and significant cross-
linking activity was observed at 160 C, even when phenolic
and nitroxyl radical traps were compounded into the mate-
rial. However, purification by repeated dissolution/precipita-
tion removed residual tetrabutylammonium bromide and
improved stability such that additive-free samples of IIR-AA
and IIR-VBA could be stored for weeks at room temperature.
The instability of IIR-MBA could not be overcome, and sam-
ples of this particular macromonomer were analyzed within
hours of their preparation.
Note that IIR degradation dynamics and yields are affected
by the presence of a polymerizable monomer. Figure 1 also
contains a plot of the storage modulus for an IIR compound
containing DCP and a small amount of styrene. The presence
of unbound monomer did not eliminate chain scission, but
reduced its extent, presumably by shifting radical popula-
tions away from primary alkyl macroradicals toward benzylic
radical intermediates. By extension, polymerizable functional
groups bound to macromonomer derivatives of BIIR may not
eliminate backbone degradation, but they can affect the bal-
ance of crosslinking and scission that dictate cure extents.
Macromonomer Crosslinking Dynamics
Figure 2 provides plots of G0 versus time for macromonomer
samples mixed with 18 mmole DCP / gramꢃpolymer and
ꢀ
heated to 160 C. As described above, the half-life of DCP at
160 ꢀC is 5.4 min and, as such, the initiator reaches 98%
conversion within the 30 min timescale of the plotted data.
Note that peroxide cures of saturated polyolefins such as
polyethylene are stoichiometric processes, and initiator ther-
molysis is the rate limiting step.34 As a result, crosslinking
dynamics of saturated polymers mirror the first-order
kinetics of initiator decomposition. The macromonomer data
Synthesis of IIR-Derived Macromonomers
The macromonomers depicted in Scheme 1 were prepared
by nucleophilic displacement of bromide from BIIR by the
required carboxylate salt.26,27 These esterifications are com-
plicated somewhat by the insolubility of alkali metal carbox-
ylate salts in solvents that dissolve the polymer. Phase-
presented in Figure
2 demonstrate altogether different
behaviour, in that cure rates are not linearly proportional to
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
5

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
IIR.36,37 As such, residual unbrominated isoprene isomers in
BIIR could copolymerize with polymer-bound maleimide
functionality to heighten cure activity. It is difficult to place
the reactivity of IIR-VImBr into proper context, since reports
of the polymerization of vinylimidazolium ionic liquids38–40
have not been accompanied by rate data.
The three acrylic macromonomers, IIR-AA, IIR-MAA, and IIR-
DDI crosslinked more slowly and to a more moderate extent
than those described above. The superior reactivity of IIR-AA
over IIR-MAA is consistent with reported propagation rate
constants,14 but the activity of the itaconate diester, IIR-DDI,
is higher than expectations based on this data set. Although
reported kp values at 60ꢀ for itaconate esters are an order of
magnitude less than those of methacrylates, at the high tem-
perature used in macromonomer crosslinking, the sterically
hindered itaconate derivative (IIR-DDI) is no less reactive
than its methacrylate analog. The three remaining macromo-
nomers, IIR-AlImBr, IIR-DDMA, and IIR-UA, were relatively
unreactive, providing a combination of low crosslinking rates
and cure extents. Nonetheless, the unactivated a-olefin func-
tionality within IIR-UA counteracted polymer backbone
degradation.
FIGURE 2 Dynamics of peroxide-initiated macromonomer
crosslinking ([DCP] 5 18 lmol/g; T 5 160 ꢀC).
Further analysis of the cure dynamics data revealed a rela-
tionship between the crosslink density provided by a macro-
monomer, as measured by the difference in storage modulus
(DG0), and the maximum rate of crosslinking (dG0/dtjmax).
The latter is determined by taking the numerical derivative
of G0 versus time, and reporting the highest rate observed
throughout the curing reaction. Figure 3 presents a semi-log
plot of DG0 against dG0/dtjmax that demonstrates the correla-
tion between kinetic reactivity and cure extent. Since all
macromonomers contained 0.15 mmol of reactive functional-
ity per gram of polymer, they had the same potential to
affect crosslinking. Nevertheless, IIR-VBA produced a higher
crosslink density than did IIR-AA.
initiator decomposition rates; they vary considerably
depending on polymerizable group structure.
The vinyl benzoate ester (IIR-VBA) was the most reactive,
progressing from an initial storage modulus of 57 kPa to a
maximum of 440 kPa (DG0 5 G0max – G0min 5 383 kPa) within
just one initiator half-life. Cure reversion was observed
beyond the 5.2 min mark of the process, at which point radi-
cal scission became the dominant molecular weight altering
reaction. This postcure loss of storage modulus occurs when
macromonomer conversion to crosslinks is no longer com-
petitive with polymer backbone degradation, a shift that
likely coincides with the consumption of pendant monomer
groups.7
The storage modulus data plotted in Figure 2 show that IIR-
VBA, IIR-MBA and IIR-VImBr provide similar cure dynamics,
as they crosslink rapidly to full monomer conversion before
undergoing significant postcure reversion. As noted by one
reviewer, potential differences in the reactivity of these three
materials may be masked by diffusion limitations imposed
on monomer functionality by the high viscosity matrix.
Nonetheless, the results are somewhat surprising, given that
propagation rate constants for acrylates at moderate temper-
atures are greater than those generally reported for styrenics
and maleimides. The exceptional reactivity of IIR-VBA may
be due to an electron withdrawing effect of the para-ester
group,14 since the propagation rate constant for dodecyl-4-
vinyl benzoate is known to exceed that of styrene.35
The reactivity of IIR-MBA may also benefit from an inductive
effect of the N-aryl substituent, which would activate the
moderate homopolymerizability of maleimide functionality.
Note, however, that maleimides copolymerize readily with
electron-rich olefins and are effective co-curing agents for
FIGURE 3 Cure yield versus maximum cure rate ([DCP] 5
18 lmole/g; T 5 160 ꢀC; t 5 60 min).
6
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
for the IIR-UA system, which failed to produce a measurable
increase in G0, irrespective of the peroxide loading used.
Conversely, if crosslinking is favoured, then it will generate
highly branched architectures whose molecular weight is rel-
atively insensitive to chain scission. The storage modulus
data illustrated in Figure 4(c) shows how IIR-VBA requires
relatively few initiator-derived radicals to generate a highly
elastic covalent polymer network. Once established, this net-
work is more robust toward chain cleavage than a linear
polymeric material. This phenomenon has been described in
studies of biomaterial degradation, where a polymer’s mass
loss due to enzymatic,42–44 hydrolytic,38,45 or thermal degra-
dation46 is strongly dependent on crosslink density, and that
degradation is much slower in a crosslinked polymer com-
pared to its linear parent material.41
Hoffmann has presented a theoretical framework that cap-
tures the effect of chain architecture on a material’s suscepti-
bility to cleavage.47 When a linear chain of molecular weight
MWo is subjected to a fixed number of random scission
events (Z), the average molecular weight (MW) of the prod-
uct can be expressed as MW 5 MWo/(Z-1). For example, a
linear polymer chain incurring two chain scission events will
have lost two thirds of its molecular weight.48 In contrast, a
scission event only affects the molecular weight of a cross-
linked polymer when it frees a chain segment from the net-
work. The number of effective scission events (Zeff) is
limited to those affecting chain segments that have just one
point of attachment to the network, since scission between
two crosslink nodes has no effect. It stands to reason that
Zeff will be considerably less than Z for highly crosslinked
polymer architectures.
In the context of IIR-based macromonomers, the most reac-
tive materials quickly establish polymer networks that are
more resistant to degradation than their linear parent mate-
rials. Therefore a material such as IIR-VBA can generate a
high storage modulus at the 5 min mark of a DCP cure,
yielding a covalent network that resists large-scale degrada-
tion over the next 15 min, even though over 50% of the per-
oxide remains active.
FIGURE 4 Crosslinking of (a) IIR-UA, (b) IIR-MAA, and (c) IIR-
VBA with varying peroxide loadings.
One potential cause of this reactivity/yield connection is
insufficient radical initiation. For example, a moderately reac-
tive material such as IIR-MAA may need more than 18 mmole
ꢀ
DCP/g to reach full monomer conversion at 160 C. However,
Delayed-Onset Macromonomer Cures
the data plotted in Figure 4(b) show that doubling the per-
oxide concentration to 36 mmole DCP/g had little effect on
the ultimate storage modulus, meaning that the ultimate
cure extent of IIR-MMA is not the result of an “initiator-
starved” polymerization, but a by product of its low kinetic
reactivity.
The macromonomer structure / reactivity studies described
above have shown that high cure yields are only obtained
using highly reactive polymerizable functionality. However,
extreme crosslinking rates in the early stages of a peroxide
cure, a phenomenon known as scorch, can render the poly-
mer thermoset before it can be molded into the desired
shape. One means of delaying the onset of polymer crosslink-
ing is to include a radical trap in the cure formulation to
quench radical populations in the initial phase of the cure,
thereby suppressing macro-radical combination. Nitroxyls
such as TEMPO (Sch. 3) can be particularly effective, as they
react with carbon-centred radicals at the diffusion limit of
bimolecular reaction velocities49 to yield spin-paired alkoxy-
amines50 without trapping alkoxyl radicals generated by the
peroxide initiator,51 or affecting the rate of peroxide
We propose that the observed relationship between cure
rate and cure yield arises from a competition between inter-
dependent crosslinking and degradation reactions. Note that
if polymer degradation is kinetically favoured, then it will
generate small chain fragments whose low molecular weight
makes them more difficult to bring up to, and beyond, the
gel point.41 As such, no matter how much initiator is charged
to an inactive macromonomer, it will not generate a tightly
crosslinked thermoset. This is demonstrated in Figure 4(a)
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
7

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
to consume all the nitroxyl charged to the formulation, as
opposed to the observed value of 4.9 min. The constant stor-
age modulus proves that TEMPO is an efficient radical trap
for polymer macroradicals as well as the methyl radicals
generated by cumyloxyl fragmentation,55 but the extended
induction time suggests that nitroxyl is regenerated during
the induction period. The result is an effective trapping ratio
greater than that calculated from initial reagent
concentrations.
Information regarding alkoxyamine instability at 160 ꢀC is
lacking, but sufficient literature exists to implicate it as a
source of TEMPO regeneration. Disproportionation of ben-
zylic alkoxyamines is well documented,56–58 and in the pres-
ent context, could yield a disubstituted styrenic adduct and
TEMPOH, the latter providing TEMPO by H-atom transfer to
initiator-derived radicals and/or by air oxidation. However,
an early report by Li et al. describes the decomposition of 1-
(2,2,6,6-tetramethylpiperidinyloxy)21-phenylethane over the
course of several hours at 140 ꢀC,59 and similar reaction
timescales were reported by Ananchenko and Fischer at 120
ꢀC.60 This suggests that alkoxyamine disproportionation
could only affect a 5 min induction period if it proceeded
much more rapidly at 160 ꢀC. An alternate decomposition
pathway was recently proposed by Gryn’ova et al. wherein
b-hydrogen atom abstraction from the alkoxyamine leads to
a ketone and aminyl radical,61 the latter producing TEMPO
via several established pathways, including the Denisov
cycle.62 Alkoxyamine breakdown through this process is
radical-initiated, with dynamics controlled by the rate of per-
oxide thermolysis. As such, it may be operative over the
timescale of our IIR-VBA experiments.
SCHEME 3 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO) and
its acrylated analog (AOTEMPO).
thermolysis.52 This method has recently been explored for
the scorch protection of polyolefins21 and vinyl-
functionalized polydimethylsiloxane.53
The data presented in Figure 5 demonstrate the ability of
TEMPO, charged to the formulation at a concentration of
0.25 molar equivalents relative to DCP-derived cumyloxyl
radicals, to suppress IIR-VBA crosslinking for 4.9 min. Unlike
controlled radical polymerizations involving nitroxyls and
styrenic monomers, our intent is not to generate pseudo-
living conditions, but to trap one quarter of the initiator-
derived radicals irreversibly. Ideally, excess nitroxyl relative
to the instantaneous radical population suppresses crosslink-
ing entirely by producing stable alkoxyamines. This condition
should persist until the entire 0.25 equivalents of TEMPO
are consumed, at which point the residual peroxide should
initiate macromonomer crosslinking unabated. At first
glance, the IIR-VBA 1TEMPO system appears to approach
this ideal, in that a stable G0 is maintained for a distinct
induction period, after which rapid curing is observed to a
Although questions remain regarding mechanisms of alkoxy-
amine decomposition,63,64 implications for
a
nitroxyl-
high ultimate modulus. However,
a closer examination
mediated IIR-VBA cure may include a prolonged induction
reveals two significant deviations from ideality – the induc-
tion period is longer than anticipated, and the cure extent
falls short of expectations.
Consider the length of the induction period. Given that
TEMPO does not affect the first-order decomposition of a
dialkyl peroxide,52 the time needed to consume the nitroxyl
should be a function of the initiator half-live (t1/2) and the
trapping ratio, which is defined as the initial nitroxyl concen-
tration divided by the total number of cumyloxyl radicals
generated by DCP thermolysis. Previous studies of delayed-
onset cures of polyethylene21,54 and vinyl-functionalized pol-
ydimethylsiloxane53 have shown that the induction phase of
an ideal TEMPO-mediated cure abides by the following
equation.
ꢀ
ꢁ
t1=2
½Nitroxylꢄ0
2½DCPꢄ0
t
ind5
ln 12
lnð1=2Þ
ꢀ
ꢁ
5:4 min
ln ð1=2Þ
9lmol=g
2ð18lmol=gÞ
5
ln 12
52:2 min
FIGURE 5 Dynamics DCP-initiated cure formulations containing
TEMPO and AOTEMPO (a. LLDPE; b. IIRVBA; trapping
ratio 5 0.25).
Since the trapping ratio ([TEMPO]o/2[DCP]o) used to gener-
ate Figure 5 was 0.25, it should have taken DCP just 2.2 min
8
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
9 V. Percec, J. H. Wang. J. Macromol. Sci. A. 2006, 28, 221–231.
period and the conversion of vinyl benzoate groups into
ketones or less reactive 1,2-disubstituted styrenic functionality.
A simple means of counteracting a loss of macromonomer con-
tent uses an acrylated nitroxyl, AOTEMPO (Sch. 3). Trapping of
primary alkyl macroradicals generated as a result of IIR back-
bone activation yields acrylate macromonomer functionality,
whose oligomerization can increase cure extent.21,54 Figure 5
demonstrates the efficacy of AOTEMPO, which provides the
equivalent induction period of TEMPO while raising the ulti-
mate storage modulus to that provided by DCP alone.
AOTEMPO provides a further advantage, in that methyl alkoxy-
amines created by nitroxyl trapping of cumyloxyl-derived
methyl radicals (Sch. 2) can copolymerize with macromonomer
functionality, thereby rendering them polymer-bound.
10 J. Ryan, F. Aldabbagh, P. B. Zetterlund, B. Yamada. React.
Funct. Polym. 2008, 68, 692–700.
11 C. E. Oxley, G. J. Wilson. Rubber Chem. Technol. 1969, 42,
1147–1154.
12 H. C. Wang, J. V. Fusco, P. Hous. Rubber World 1994, 211,
37–41.
13 K. Takenaka, M. Suzuki, H. Takeshita, M. Miya, T. Shiomi, K.
Tamamitsu, T. Konda. J. Polym. Sci. Part A: Polym. Chem.
2011, 50, 659–664.
14 K. Matyjaszewski, T. P. Davis. Handbook of Radical Polymer-
ization; Wiley: Hoboken, NJ, 2002; p 534.
15 L. Wojnarovits, E. Takacs. Radiat. Phys. Chem. 1999, 55, 639–644.
16 J. D. Druliner. J. Phys. Org. Chem. 1995, 8, 316–324.
17 T. Otsu, N. Toyoda. Makromol. Chem. Rapid Commun.
1981, 2, 79–81.
CONCLUSIONS
18 C. E. Schildknecht. Allyl Compounds and Their Polymers
(Including Polyolefins); Wiley: New York, 1973.
Macromonomer derivatives of BIIR undergo simultaneous
crosslinking and degradation when activated by DCP at 160
ꢀC, with the competitive balance dictated by the reactivity of
the oligomerizable group. Highly reactive styrenic, vinyl imi-
dazolium, and maleimidic functionalities rapidly produce a
covalent polymer network that is resistant to chain scission,
thereby generating exceptional ultimate crosslink densities.
Acrylic-based macromonomers provided moderate cure rates
and yields, whereas relatively unreactive maleate diester and
allylic functionalized materials proved incapable of generat-
ing high crosslink yields, irrespective of the amount of perox-
ide used, due to the predominance of chain cleavage on
polymer molecular weight.
19 W. Wu, J. S. Parent, S. S. Sengupta, B. I. Chaudhary.
J. Polym. Sci. Part A: Polym. Chem. 2009, 47, 6661–6670.
20 R. A. Prasath, S. Ramakrishnan. J. Polym. Sci. Part A:
Polym. Chem. 2005, 43, 3257–3267.
21 D. K. Hyslop, J. S. Parent. Macromolecules 2012, 45, 8147–
8154.
22 G. E. Garrett, E. Mueller, D. A. Pratt, J. S. Parent. Macromo-
lecules 2014, 47, 544–551.
23 R. Brown. Physical Testing of Rubber; Springer: New York,
2006.
24 F. C. Auluck, D. S. Kothari. J. Chem. Phys. 1943, 11, 387–
392.
25 .Technical Bulletin, Atofina Chemicals 2001.
26 S.A. Guillen-Castellanos, J. S. Parent, R. A. Whitney. Macro-
molecules 2006, 39, 2514–2520.
Complications associated with the high initial crosslinking
rate of IIR-VBA can be overcome using nitroxyls to quench
radical activity in the early stages of the cure. Compared
with ideal delayed-onset behaviour observed for the cross-
linking of saturated polymers, the combination of IIR-VBA
and TEMPO gives prolonged induction times and reduced
cure extents. The latter can be obviated through the use of
AOTEMPO to recover macromonomer functionality that is
lost to undesired vinyl benzoate trapping reactions.
27 J. S. Parent, S. Malmberg, J. K. McLean, R. A. Whitney. Eur.
Polym. J. 2010, 46, 702–708.
28 J. K. McLean, S. A. Guillen-Castellanos, J. S. Parent, R. A.
Whitney. Ind. Eng. Chem. Res. 2009, 48, 10759–10764.
29 J.R. McNeish, J.S. Parent, R.A. Whitney. Canadian J. Chem.
2013, 91, 420–427.
30 J.S. Parent, D. Thom, G. White, R. A. Whitney, W. Hopkins.
J. Polym. Sci. Part A: Polym. Chem. 2001, 39, 2019–2026.
31 S. M. Malmberg, J. S. Parent, D. A. Pratt, J. K. McLean, R.
A. Whitney. Macromolecules 2010, 43, 8456–8461.
32 S. A. Canary, M. P. Stevens. J. Polym. Sci. Part A: Polym.
Chem. 1992, 30, 1755–1760.
REFERENCES
1 T. Shaffer, A. Tsou, R. Webb. In Kirk-Othmer Encyclopedia of
Chemical Technology; Kirk-Othmer, Ed.; John Wiley and Sons,
2007; Vol. 4.
33 D. J. T. Hill, L. Y. Shao, P. J. Pomery, A. K. Whittaker. Poly-
mer 2001, 42, 4791–4802.
34 J. Liu, W. Yu, C. Zhao, C. Zhou. Polymer 2007, 48, 2882–2891.
2 A. Y. Coran. In Science and Technology of Rubber, 3rd ed.;
Elsevier Academic Press: Boston, 2005; Chapter 7.
35 H. S. Bisht, S. S. Ray, D. Pandey, C. D. Sharma, A. K.
Chatterjee. J. Polym. Sci. Part A: Polym. Chem. 2002, 40, 1818–
1830.
3 R. T. Morrissey. Ind. Eng. Chem. 1955, 47, 1562–1569.
4 J. S. Parent, G. D. F. White, D. Thom, R. A. Whitney, W.
Hopkins. J. Polym. Sci. Part A: Polym. Chem. 2003, 41, 1915–
1926.
36 K. V. S. Shanmugam, J. S. Parent, R. A. Whitney. Eur.
Polym. J. 2012, 48, 841–849.
37 K. V. S. Shanmugam, J. S. Parent, R. A. Whitney. Ind. Eng.
Chem. Res. 2012, 51, 8957–8965.
5 P. R. Dluzneski. Rubber Chem. Technol. 2001, 74, 451–493.
6 L. D. Loan. J. Polym. Sci. Part A: Polym. Chem. 2003, 2,
38 M. Hirao, K. Ito, H. Ohno. Electochim. Acta 2000, 45, 1291–
2127–2134.
1294.
7 S. Xiao, J. S. Parent, R. A. Whitney, L. K. Knight. J. Polym.
Sci. Part A: Polym. Chem. 2010, 48, 4691–4696.
39 J. Yuan, M. Antonietti. Macromolecules 2011, 44, 744–750.
40 H. Mori, M. Yahagi, T. Endo. Macromolecules 2009, 42,
8 K. Ito. Prog. Polym. Sci. 1998, 23, 581–620.
8082–8092.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
9

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
41 J. S. Parent, A. Bodsworth, S. S. Sengupta, M.
Kontopoulou, B. I. Chaudhary, D. Poche, S. Cousteux. Polymer
2009, 50, 85–94.
52 S. M. Kabun, A. L. Buchachenko, Bull. Acad. Sci. USSR Div.
Chem. Sci. 1966, 15, 1430–1431.
53 S. Mani, P. Cassagnau, M. Bousmina, P. Chaumont. Macro-
molecules 2009, 42, 8460–8467.
42 O. Jeon, S. J. Song, K.-J. Lee, M. H. Park, S.-H. Lee, S. K.
Hahn, S. Kim, B.-S. Kim. Carbohydr. Polym. 2007, 70, 251–257.
54 D. K. Hyslop, J. S. Parent. Polymer 2013, 54, 84–89.
43 G. Imparato, F. Urciuolo, C. Casale, P. A. Netti. Biomaterials
2013, 34, 7851–7861.
55 D. V. Avila, C. E.Brown, K. U. Ingold, J. Lusztyk. J. Am.
Chem. Soc. 1993, 115, 466–470.
44 G. J. M. de Koning, H. M. M. van Bilsen, P. J. Lemstra, W.
Hazenberg, B. Witholt, H. Preusting, J. G. van der Galien, A.
Schirmer, D. Jendrossek. Polymer 1994, 35, 2090–2097.
56 C. J. Hawker, A. W. Bosman, E. Harth. Chem. Rev. 2001,
101, 3661–3688.
57 K. Ohno, Y. Tsujii, T. Fukuda. Macromolecules 1997, 30,
45 K. A. George, T. V. Chirila, E. Wentrup-Byrne. Polym.
Degrad. Stab. 2012, 97, 964–971.
2503–2506.
58 D. Greszta, K. Matyjaszewski. Macromolecules 1996, 29,
46 V. V. Krongauz. Termochim. Acta 2010, 503–504, 70–84.
47 M. Hoffmann. Makromol. Chem. 1982, 183, 2191–2211.
7661–7670.
59 I. Li, B. A. Howell, K. Matyjaszewski, T. Shigemoto, P. B.
Smith, D. B. Priddy. Macromolecules 1995, 28, 6692–6693.
48 W. Schnabel. In Polymer Degradation: Principles and Practi-
cal Applications; Hanser International: Muenchen, 1981; p 21.
60 G. S. Ananchenko, H. Fischer. J. Polym. Sci. Part A: Polym.
Chem. 2001, 39, 3604–3621.
49 J. Chateauneuf, J. Lusztyk, K. U. Ingold. J. Org. Chem. 1987,
53, 1629–1632.
61 G. Gryn’ova, K. U. Ingold, M. L. Coote J. Am. Chem. Soc.
2012, 134, 12979–12988.
50 B. I. Chaudhary, L. Chopin, J. Klier. Polym. Eng. Sci. 2006,
47, 50–61.
62 N. S. Allen. Chem. Soc. Rev. 1986, 15, 373–404.
63 F. Gugumus. Polym. Degrad. Stab. 1993, 40, 167–215.
64 J. Pospisil. Adv. Polym. Sci. 1995, 124, 87–189.
51 A. L. J. Beckwith, V. W. Bowry, M. Graeme. J. Org. Chem.
1988, 53, 1632–1641.
10
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000