Imidazolium bromide derivatives of poly(isobutylene-co-isoprene): A new class of elastomeric ionomers

Article abstract of DOI:10.1016/j.polymer.2011.10.021

New imidazolium bromide ionomers are prepared by halide displacement from brominated poly(isobutylene-co-isoprene) (BIIR) with a variety of imidazole-based nucleophiles. N-alkylation of butylimidazole (BuIm) by the allylic halide functionality within BIIR provides stable imidazolium bromide functionality in high yields. Characterization of the resulting thermoformable ionomer by dilute solution viscosity and solid-state rheology shows the effects of ion-pair aggregation on material properties over a range of ionomer composition. Knowledge of BuIm alkylations is extended to produce thermoset ionomer derivatives through solvent-free reactions of BIIR with 1,1′(1,4-butanediyl)bis(imidazole) and imidazole, whose physical properties stem from both a covalent network and a network of ion-pair aggregates. A novel method of preparing these thermosets is described, wherein a poly(isobutylene-co-isoprene) derivative bearing pendant imidazole functionality is prepared, and heated with BIIR parent material to yield covalent crosslinks comprised of dialkylated imidazole moieties.

Full text of DOI:10.1016/j.polymer.2011.10.021

Polymer 52 (2011) 5410e5418
Contents lists available at SciVerse ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Imidazolium bromide derivatives of poly(isobutylene-co-isoprene): A new class
of elastomeric ionomers
J. Scott Parent a , Anthony M.J. Porter , Monika R. Kleczek , Ralph A. Whitney
,
a
a
b
*
a
Department of Chemical Engineering, Queen’s University, Kingston, ON, Canada K7L 3N6
Department of Chemistry, Queen’s University, Kingston, ON, Canada K7L 3N6
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 August 2011
Received in revised form
New imidazolium bromide ionomers are prepared by halide displacement from brominated
poly(isobutylene-co-isoprene) (BIIR) with a variety of imidazole-based nucleophiles. N-alkylation of
butylimidazole (BuIm) by the allylic halide functionality within BIIR provides stable imidazolium
bromide functionality in high yields. Characterization of the resulting thermoformable ionomer by dilute
solution viscosity and solid-state rheology shows the effects of ion-pair aggregation on material prop-
erties over a range of ionomer composition. Knowledge of BuIm alkylations is extended to produce
5
October 2011
Accepted 8 October 2011
Available online 14 October 2011
0
thermoset ionomer derivatives through solvent-free reactions of BIIR with 1,1 (1,4-butanediyl)bis(imi-
Keywords:
dazole) and imidazole, whose physical properties stem from both a covalent network and a network of
ion-pair aggregates. A novel method of preparing these thermosets is described, wherein a poly(-
isobutylene-co-isoprene) derivative bearing pendant imidazole functionality is prepared, and heated
with BIIR parent material to yield covalent crosslinks comprised of dialkylated imidazole moieties.
Ó 2011 Elsevier Ltd. All rights reserved.
Ionomer
Elastomer
Polymer modification
1. Introduction
ionomers, and the preparation of thermoset ionomers using single-
step polymer compounding techniques. While previous research has
The introduction of small amounts of ionic functionality to non-
polar polymers yields ionomer derivatives that provide mechanical
and adhesive properties that are superior to their parent materials
focused on the synthesis and characterization of chemically inert
ionomers, the potential to introduce ion-pairs that bear reactive
functionality has been largely overlooked. Unfortunately, conven-
tional triarylphosphines and tertiary amines are ill-suited for this
application. The air-instability of most phosphines, coupled with the
challenges in preparing functional phosphine nucleophiles [13],
limits the practical range of P-alkylation chemistry to triphenyl-
[1,2]. Ionomers bearing carboxylate and sulfonate functionality have
been studied extensively [3,4], but cationic ionomers have also
attracted recent attention [5], owing to their activity against a range
of bacteria and fungi [6] as well as their resistance to fouling by
marine mollusks [7]. We are particularly interested in derivatives of
isobutylene-rich elastomers, since these materials provide excep-
tional impermeability, oxidative stability, and vibration dampening
characteristics [8]. A simple route to cationic ionomer elastomers
involves nucleophilic displacement of halide from halogenated
starting materials such as brominated poly(isobutylene-co-
paramethylstyrene) (BIMS) and brominated poly(isobutylene-co-
isoprene) (BIIR). Solvent-free reactions of these halogenated poly-
mers with PPh and tertiary amines are known to provide corre-
3
sponding quaternary phosphonium [9,10] and ammonium [11,12]
bromide ionomers in good yield.
This manuscript describes two advances to this ionomer chem-
istry e the development of a new class of imidazolium bromide
3
phosphine (PPh ). On the other hand, tertiary amines containing
reactive functional groups are available, but the vulnerability of
tetraalkylammonium bromides to dealkylation is troublesome [14].
Excess amine is needed to push BIIR reactions to high conversion,
and the resulting ammonium bromide salt is unstable to amine
displacement by bromide or other nucleophiles [15].
This work is concerned with imidazole-derived ionomers that
support a wider range of ionomer chemistry than ammonium and
phosphonium analogues. A surge in the ionic liquid field has fueled
interest in functional imidazoles, and has clearly demonstrated the
superiour stability of their corresponding imidazolium bromide
salts [16,17]. Our studies began with the synthesis and character-
ization of thermoformable imidazolium bromide materials
produced through N-alkylation of 1-butylimidazole (BuIm) by the
allylic bromide functionality within BIIR (Scheme 1). An under-
standing of reaction dynamics is used to prepare materials with
different ion-pair contents for subsequent studies of structure-
*
Corresponding author. Tel.: þ1 613 533 6266; fax: þ1 613 533 6637.
E-mail address: parent@chee.queensu.ca (J.S. Parent).
0
032-3861/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2011.10.021

J.S. Parent et al. / Polymer 52 (2011) 5410e5418
5411
Bu
cat. Br
S 2'
N
N
Bu
N
N
Br
Br
Br
N
S 2
N
2a
1
b
Br
1
a
Br
Bu
Br
N
Br
Bu
N
N
Br
N
1
c
SN2
2b
Scheme 1. N-alkylation of BuIm by 1.
property relationships. This basic knowledge serves as a precursor
to future investigations of functional ionomers, and provides
a foundation for developing thermoset imidazolium ionomers.
The need for thermoset ionomer chemistry stems from the
susceptibility of uncross-linked elastomeric ionomers to creep and
stress relaxation [18,19]. It is well known that ion-pairs are poorly
solvated by a polymer backbone of low dielectric constant, and self-
assembly of ionic functionality is thermodynamically favorable
polydispersity ¼ 1.5) containing 0.15 mmol/g of allylic bromide
functionality, was used as supplied by LANXESS Inc. (Sarnia, Ontario).
2.2. Synthesis of 1-n-butyl-3-[(2E,Z)-6,6-dimethyl-2-
neopentylhept-2-enyl] imidazolium bromide (2a,b)
BPMN (0.01 mL, 0.9 g, 42.0
48 h at 85 C to isomerize 1a to a 60:40 ratio of 1b:1c. The resulting
m
mol) was heated in toluene-d
8
for
ꢀ
[
20]. The resulting aggregation of ion-pairs establishes a non-
solution was then charged with 1-butylimidazole (0.010 mL,
ꢀ
covalent network of polymer chains, whose dynamic mechanical
properties approach those of conventional covalent thermosets
that are comprised of carbonecarbon and/or sulfide cross-links.
However, the lability of an ionic network leads to a poor response
to static loads, and thermoformable ionomers do not meet the
standards of engineering materials. Thermoset ionomers
comprised of covalent cross-link networks and ionic functionality
are required, and a direct route to their preparation is highly
desirable.
0.0095 g, 76.5
mmol) and heated for 10 h at 85 C to give 2a,b in
quantitative yield. Solvent and residual reagents were removed by
ꢀ
Kugelrohr distillation at 65 C, 80 Pa. High-resolution MS analysis:
þ
1
required for C21
(CDCl ):
0.8e2.2 (m, 2 ꢁ eC(CH
(t, NeCH e(CH eCH ), Found for Z-isomer 2a:
N¼CHeN), 7.29 (m, 1H, NCHCHN), 7.08 (d, 1H, NCHCHN), 5.70
(t, HeC¼), 4.95 (s,¼CeCH eN). Found for E-isomer 2b: 10.59
(s, 1H, N¼CHeN), 7.30 (m, 1H, NCHCHN) 7.12 (d, 1H, NCHCHN), 5.65
H
39
N
2
m/z 319.3113; found m/z 319.3089. H NMR
, 5 ꢁ eCH e, eCH ), 4.41
10.67 (s, 1H,
3
d
3
)
3
2
3
2
2
)
2
3
d
2
d
1
The second half of this paper describes an extension of BuIm
alkylation to prepare imidazolium thermosets using conventional
(t, HeC ¼ ), 4.86 (s, ¼ CeCH
2
eN). 2D ROESY H NMR was used to
differentiate E and Z isomers.
elastomer processing methods. Products derived from solid-state
0
reactions of BIIR
þ
1,1 (1,4-butanediyl)bis(imidazole) and
2.3. Synthesis of IIR-BuImBr
BIIR þ imidazole are characterized using model compounds before
studying their cure dynamics using rheological measurements. We
conclude with the preparation of an isobutylene-rich elastomer
bearing pendant imidazole functionality, which is shown to provide
a novel thermoset ionomer when heated with its halogenated
parent material.
BIIR (10.0 g, 1.5 mmol allylic bromide) and 1-butylimidazole
(1.12 g, 9.0 mmol) were dissolved in toluene (104 mL) and heated
ꢀ
to 100 ꢂ 2 C for 6 h under nitrogen. Aliquots (w0.5 mL) withdrawn
at intervals were precipitated from excess acetone to isolate the
polymeric reaction product, which was dried under vacuum and
1
characterized by H NMR spectroscopy. Samples with different
2
. Experimental section
imidazolium bromide contents were prepared by this method
using Bu
reaction times. For example, an IIR-BuImBr sample containing
.05 mmol/g of imidazolium bromide was prepared by stopping N-
alkylation after 3 h by adding excess Bu N acetate (0.678 g,
4
N acetate to quench residual electrophile at specific
2
.1. Materials
0
Brominated 2,2,4,8,8-pentamethyl-4-nonene (BPMN, 1) was
4
prepared as described previously [21]. The following reagents were
2.25 mmol) and heating for a further 2 h to convert residual allylic
bromide to the corresponding acetate ester. Products were
precipitated from acetone and purified by dissolution-precipitation
(THF-acetone) prior to drying under vacuum.
used as received from SigmaeAldrich (Oakville, Ontario): 1,8-
0
bis(dimethylamino)naphthalene (Proton Sponge, 99%), 4,4 -bipyr-
idyl (99%), 1-butylimidazole (BuIM, 98%), 1,4-dibromobutane
(
99%),1,4-diazabicyclo[2.2.2]octane
(DABCO,
97%),
ethyl-
0
enebis(diphenylphosphine) (DIPHOS, 97%), 1-(2-hydroxethyl)-imid-
azole (97%), imidazole (99%), succinic anhydride (99%),
tetrabutylammonium acetate (TBA acetate, 97%), and tricaprylme-
thylammonium chloride (Aliquat 336). Brominated poly(isobu-
tylene-co-isoprene) (BIIR or BB2030,
2.4. Synthesis of 1,1 (1,4-butanediyl)bis(imidazole) (3)
NaOH (12.50 g) was dissolved in water (12.50 g) mixing with
toluene (60 mL). Imidazole (1.31 g, 19.8 mmol) and Bu
1.98 mmol) were mixed into the solution at 25 C until all solids
Ò
4
NBr (0.606 g,
ꢀ
M
w
¼
410 000 and

5412
J.S. Parent et al. / Polymer 52 (2011) 5410e5418
were dissolved prior to adding 1,4-dibromobutane (2.02 g,
of curing time using a dynamic oscillatory rheometer (Advanced
Polymer Analyzer 2000; Alpha Technologies) operating in a parallel
plate configuration with a 3 arc at a frequency of 1 Hz.
ꢀ
ꢀ
9
.32 mmol) and heating to 60 C for 16 h. Upon cooling to 25 C,
ꢀ
water (80 mL) was added to the mixture before recovering crystals
1
by vacuum filtration, giving compound 3 (1.48 g, 84% yield).
NMR (CDCl 1.76 (t, eCH eCH eCH eCH e, 4H), 3.93 (m,
NeCH eCH
H
3
)
d
2
2
2
2
2.9. Analysis
2
2
e, 4H), 6.86 (s, 2H), 7.07 (s, 2H), 7.44 (s, 2H).
mp ¼ 83e86 C [22].
ꢀ
NMR spectra were recorded with a Bruker Advance-600 spec-
1
13
trometer (600.17 MHz H, 150.92 MHz C) in deuterated chloro-
form (CDCl ) with chemical shifts referenced to tetramethylsilane.
2.5. Synthesis of 1,3-bis-[(2)-6,6-dimethyl-2-neopentylhept-2-enyl]
3
imidazolium bromide (4)
Solution viscosities were measured using Cannon-Fenske and
Ubbelohde viscometers so that the elution time was at least 200 s,
which eliminated the need for kinetic energy corrections. Data
ꢀ
BPMN (0.02 mL, 0.18 g, 0.148 mmol) was heated for 48 h at 85 C in
toluene-d in anNMR tube to getan equilibratedratioofE and Z-endo
isomers. The isomerized model compound was then reacted with
imidazole (5.05 mg, 74.2 mol) and Proton Sponge (15.0 mg,
8
acquired in the dilute solution region (below 1 g/dL) were fit to
2
Huggins equation
viscosity (dL/g), [
h
red ¼ [
h
] þ k
H
[
h
] c, where
h
red is the reduced
m
h
] is the intrinsic viscosity, c is the polymer
is the Huggins coefficient
23]. High-resolution mass spectrometry analysis was conducted
ꢀ
mmol) for 6 h at 100 C. Solvent and residual reagents were
removed by Kugelrohr distillation at 65 C, 80 Pa, yielding compound
74.2
concentration in solution (g/dL) and k
H
ꢀ
[
4
as a mixture of Z,Z, Z,E, and E,E geometric isomers. High-resolution
using a Waters/Micromass GC-T TOF instrument with electron
impact ionization.
þ
MS analysis: required for C31
H
57
N
2
m/z 457.4521; found m/z
1
4
1
(
57.4544. H NMR (CDCl
0.79 Z,Z-isomer; 10.67 E,Z-isomer; 10.55 E,E-isomer), 7.65e7.94
three multiplets, 2H, NCHCHN), 5.69 (t, HeC ¼ , Z-isomer), 5.65 (t,
HeC E-isomer), 4.99 (s, CeCH eN, Z-isomer), 4.91
s, ¼ CeCH eN, E-isomer), 0.8e2.2 (m, 4 ꢁ eC(CH , 6 ꢁ eCH e). 2D
ROESY H NMR was used to differentiate E and Z isomers.
3
):
d
10.70e10.55 (3 singlets, 1H, N¼CHeN,
3. Results and discussion
¼
,
¼
2
3.1. Alkylation of 1-butylimidazole
(
2
)
3 3
2
1
The BIIR used throughout this study was a high molecular weight
elastomer comprised of a random distribution of approximately
7 mol% isobutylene mers, 1 mol % isoprene mers, and 2 mol%
2
.6. Synthesis of 4-[2-(1H-imidazol-1-yl)ethoxy]-4-oxobutanoic
9
acid (5)
brominated isoprene mers (allylic bromide functionality). Charac-
terizing chemically modified derivatives of this polymer is compli-
cated by the low concentration of functionality within the material,
and the impossibility of isolating different reaction products for
unambiguous structural determination. Therefore, we characterized
the products of BIIR þ 1-butylimidazole (BuIm) reactions by
comparison to derivatives of brominated pentamethylnonene (1,
Scheme 1), a model compound for the allylic halide functionality in
the polymer [24,25]. Reaction of 1 with BuIm yielded 1-n-butyl-3-
1
-(2-hydroxyethyl)imidazole (1.96 g, 0.0178 mol) and succinic
anhydride (1.78 g, 0.0178 mol) were added to THF (20 mL) and
refluxed for 6 h. Off-white crystals were collected by gravity
filtration and dried overnight.
eCH
1
H
NMR (D
eCH eCOOH, 2H), 4.39 (t,
e, 2H), 4.44 (t, eNeCH eCH eO, 2H), 7.36 (s,
2
O) d 2.34 (t,
2
eCH
2
eCOOR, 2H), 2.49 (t, eCH
2
2
eNeCH
2
eCH
2
2
2
eNeCH ¼ CHeNeR, 1H) 7.46 (s, eNeCH ¼ CHeNeR, 1H), 8.65 (s,
eN¼CHeNe, 1H). High-resolution MS analysis: required for
ꢀ
[(2E,Z)-6,6-dimethyl-2-neopentylhept-2-enyl]
imidazolium
C
9
H
12
N
2
O
4
m/z 212.0797; found m/z 212.0802. mp ¼ 108e110 C.
bromide as a mixture of E,Z-isomers (2a,b), with no exomethylene
isomer analogous to 1a found amongst reaction products. This is
consistent with BIIR-derived products, which showed evidence only
of E,Z-IIR-BuImBr functionality analogous to 2ab (Fig. 1). We
conclude, therefore, that IIR-BuImBr is comprised of a random
distribution of imidazolium ion-pairs, whose concentration is
2
.7. Synthesis and cross-linking of IIR-g-imidazole
4
-[2-(1H-imidazol-1-yl)ethoxy]-4-oxobutanoic acid (0.372 g,
1
.76 mmol), Aliquat 336 (0.710 g, 1.76 mmol) and potassium
hydroxide (0.0704 g, 1.76 mmol) were mixed in toluene (5 mL) for
7 h at 25 C. The resulting mixture was added to BIIR (3.57 g,
.536 mmol) in toluene (75 mL) and heated to 100 C for 1 h, giving
ꢀ
1
0
ꢀ
1
IIR-g-imidazole.
H NMR analysis of the product confirmed
complete conversion of allylic bromide to the corresponding esters
[
(
27]. For the purposes of cross-linking analysis, IIR-g-imidazole
3.5 g) and BIIR (3.5 g) were mixed by dissolving in toluene and
precipitating in acetone. The resulting mixture was re-dissolved in
THF and precipitated into excess acetone to facilitate vacuum
drying.
2.8. Cross-linking studies
BIIR formulations for cross-linking studies were prepared in
a Haake Polylab R600 internal mixer equipped with Banbury blades
at 60 rpm for 5 min. Formulations are specified in terms of the
molar equivalents of nucleophile and other additives relative to the
0
.15 mmol of allylic bromide functionality per gram of BIIR. For
example, a cross-linking reaction with 0.5 equivalents of compound
3
3
involved mixing BIIR (40 g, 6 mmol allylic bromide) with 3 (0.58 g,
ꢀ
0
.0 mmol) in an internal batch mixer at 90 C. The storage (G ) and
Fig. 1. Downfield region of 1H NMR spectra (CDCl
(bottom).
3
) of 2a,b (top) and IIR-BuImBr
0
0
loss (G ) modulus of the formulation was monitored as a function

J.S. Parent et al. / Polymer 52 (2011) 5410e5418
5413
limited to the 0.15 mmol of allylic bromide per gram of BIIR starting
material. Note that the low concentration of imidazolium function-
ality differentiates our materials from poly(ionic liquids) derived
from imidazolium-containing monomers [26]. Hereafter we use the
term IIR-BuImBr-0.XX to indicate ionomer composition, where 0.XX
is the mmoles of imidazolium bromide functionality per gram of
ionomer.
Dialkylimidazolium bromides similar to 2a,b are reported to be
robust ionic liquids [16], unlike quaternary ammonium bromides
that can be susceptible to dealkylation [14]. Whereas BIIR reactions
with trialkylamines require excess nucleophile to push their
alkylation to full allylic bromide conversion [15], one equivalent of
BuIm was sufficient to achieve the same end. Furthermore, heating
bromide distributed in an 80:20 ratio of functionality analogous to
structures 2a:2b.
Bromide displacement from BIIR has been studied for a wide
range of nucleophiles, resulting in a good understanding of their
reaction dynamics and product distributions [28]. A general
mechanism that accounts for BuIm alkylations is presented in
Scheme 1. Although the rearrangement of Exo-Br (1a) can occur by
ionization and ion-pair collapse to give an equal distribution of E,Z-
ꢀ
BrMe isomers (1b,1c), this pathway is very slow at 100 C [21].
0
N
Therefore, allylic bromide isomerization occurs mainly by an S 2
displacement by free bromide anion, resulting in selective Exo-Br
(1a) rearrangement to Z-BrMe (1b) as opposed to E-BrMe (1c)
[29]. The origin of this selectivity has been described previously,
and is due to a preferential conformation of Exo-Br functionality
a 10 wt% solution of IIR-BuImBr-0.15 in toluene with 1.5 equivalents
ꢀ
of Bu
Bu
4
N acetate yielded no exchange product after 2 h at 100 C.
N
that is preserved during substitution [30]. Simple S 2 displacement
4
N acetate is a good nucleophile toward BIIR, giving stable allylic
of bromide by BuIm from E,Z-BrMe groups gives the observed
products. Based on this mechanism, the observed preference for Z-
imidazolium bromide (2a) over the E-isomer (2b) is the result of
selective rearrangement of the Exo-Br functionality within our BIIR
starting material.
esters in quantitative yield [27]. That IIR-BuImBr was inert toward
this reagent distinguishes it from its ammonium analogues, from
which tertiary amines are readily displaced under the same
conditions [15].
Having established the structure and stability of BuIm alkylation
products, our studies shifted to reaction dynamics. Like BIIR qua-
ternizations of PPh and NR , BuIm reaction rates are sensitive to
3 3
3.2. Properties of IIR-BuImBr
nucleophile concentrations (as expected for a bimolecular substi-
tution), leading us to accelerate the process by using 6 equivalents
BuIm relative to the allylic bromide functionality within BIIR. Fig. 2
illustrates the evolution of allylic and imidazolium bromide func-
To understand the influence of ion-pair concentration on ion-
omer properties, we prepared samples with a range of composi-
tions by repeating the preparation illustrated in Fig. 2 several times,
4
using Bu N acetate to quench residual allylic bromide at various
ꢀ
tionality for a BIIR þ BuIm reaction carried out in toluene at 100 C.
conversions. At the extremes of our composition range were IIR-
BuImBr-0.15, which contained 0.15 mmol imidazolium bromide
per gram polymer, and IIR-BuImBr-0.00, which contained no imi-
dazolium bromide functionality and 0.15 mmol acetate ester per
gram. Both materials dissolved in toluene at room temperature, and
could be filtered with coarse glass frits without material losses.
Fig. 3 shows reduced viscosity plots for different concentrations of
these materials in toluene solution.
Starting from a 90:0:10 ratio of allylic bromides analogous to
structures 1a:1b:1c, full conversion was reached in 6 h. The
product, IIR-BuImBr-0.15, contained 0.15 mmol/g of imidazolium
Measurements of dilute solution viscosity probe intra-
molecular interactions, since polymer chains are isolated within
the solvent medium. For ionomer systems, ion-pair aggregation
reduces the radius of gyration of polymer chains and, by extension,
lowers intrinsic viscosity values [31,32]. In our case, data recorded
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
2
1
0
0
.0
g/dl 1.0
IIR-BuImBr-0.00
IIR-BuImBr-0.15
0.0
3.0
6.0
9.0
Polymer Concentration (g/dl)
Fig. 3. Reduced solution viscosity as a function of IIR-BuImBr-0.00 (B) and IIR-
ꢀ
BuImBr-0.15 (A) concentrations (toluene; 27 C; Inset: expansion of low concentra-
tion region).
ꢀ
Fig. 2. Dynamics of BuIm alkylation (10 wt% BIIR in toluene; 6 eq. BuIm; 100 C).

5414
J.S. Parent et al. / Polymer 52 (2011) 5410e5418
in the very dilute region (<1 g/dl) revealed significantly lower
solution viscosities for the ionomer compared to the non-ionic
analogue. Extrapolating reduced viscosity data to zero concentra-
polymer stiffness and elasticity [18]. The data presented in Fig. 4b
show that the introduction of imidazolium bromide functionality has
a substantial effect on material properties, with intermolecular ion-
pair interactions providing a polymer chain network with improved
modulus and elasticity. However, unlike dilute solution viscosity, the
solid-state rheological properties of IIR-BuImBr are most sensitive at
tion gave intrinsic viscosity values, [h], of 0.11 dl/g for IIR-BuImBr-
0
.15, compared to 1.03 dl/g for IIR-BuImBr-0.00. This is consistent
with previous studies on BIIR-derived phosphonium ionomers,
whose ionic aggregation in solution has been explored in greater
detail by NMR relaxation measurements [12]. The semi-dilute
region of the viscosity plot is sensitive to intermolecular associa-
tions, as demonstrated by Kennedy and coworkers for sulfonated
polyisobutylene telechelic ionomers [33]. The data plotted in Fig. 3
show that IIR-BuIm-Br solution viscosity rises well above that of the
non-ionic polymer at concentrations above 4 g/dL. Clearly, the
critical entanglement concentration of IIR-BuImBr-0.15 is consid-
erably lower, illustrating the heightened strength of inter-chain
associations provided by imidazolium bromide functionality.
Further information regarding the relationship between poly-
mer properties and ion-pair content is presented in Fig. 4. The effect
0
the lower end of our ion-pair content range. A 60% increase in G was
observed upon introducing just 0.05 mmol/g of ionic functionality to
0
the elastomer backbone. However, incremental improvements of G
diminished beyond this ion-pair content, with only small improve-
ments observed beyond 0.11 mmol/g. A similar plateau in mechanical
properties has been observed for sulfonate derivatives of
poly(isobutylene-co-isoprene), for which tensile strength and melt
viscosity leveled off beyond 0.02 milliequivalents of ionic function-
ality per gram of elastomer [35].
The new thermoformable imidazolium bromide ionomers
described above are interesting in their own right, but do not afford
material properties or chemical stability that differ significantly
from existing triphenylphosphonium bromide materials. The
important difference is that imidazole-based nucleophiles bearing
additional functionality are easily prepared and handled, whereas
the synthesis of functional triarylphosphines is considerably more
difficult. We are currently preparing ionomers bearing N-vinyl
substituents, trialkoxysilane groups and other reactive function-
ality, and characterizing their utility in peroxide and moisture
curing applications. In keeping with our current focus on simple
alkylimidzolium systems, we now describe new chemistry for
preparing thermoset ionomers.
of imidazolium bromide content on [h] is particularly interesting, in
that intrinsic viscosities did not vary linearly between our extreme
compositions. Rather, ion-pair contents less than 0.05 mmol/g had
a small effect on [h], while those above 0.11 mmol/g produced
substantial viscosity reductions. This suggests that a threshold
concentration of ion-pairs exists, below which intramolecular
associations affect hydrodynamic radii only marginally, and above
which the aggregation of imidazolium bromide functionality is
highly influential. It stands to reason that a third domain should
exist at higher ion-pair contents, where [h] would approach
a small, non-zero plateau. This presumes that the polymer will
always contribute to solution viscosity. Unfortunately, our starting
material contained only 0.15 mmol of electrophile per gram of BIIR,
which precluded the preparation of materials in the required
composition range.
0
3.3. Alkylation of 1,1 (1,4-butanediyl)bis(imidazole)
All commercial rubber products are thermoset articles
comprised of sulfidic cross-links if dynamic properties are valued,
and carbonecarbon cross-links if thermochemical stability is para-
mount [36]. Conventional cross-linking processes involve mixing
the elastomer with filler and curatives under solvent-free condi-
tions, and compression molding the resulting compound at high
temperature, thereby shaping and curing the material simulta-
neously. As noted above, uncured ionomers can provide dynamic
moduli on the same order of magnitude as covalent thermosets.
However, stress relaxation of labile ion-pair aggregates makes them
unsuitable as engineering materials, owing to creep and compres-
sion set when exposed to a static load for prolonged periods [37].
Scheme 2 illustrates a new approach to ionomer synthesis, wherein
a bis-imidazolium thermoset comprised of covalent cross-links and
The influence of ion-pair aggregation on the solid-state rheolog-
ical properties of ionomers has been reviewed extensively [20,34],
and dynamic oscillatory measurements of solid-state storage
0
modulus (G ) and loss tangent (tan
d) are standard measures of
a
0
ionic aggregates is prepared directly from 1,1 (1,4-butanediyl)bis(i-
midazole), 3, under solvent-free conditions. By introducing imida-
zolium bromide functionality in the form of covalent cross-links, the
engineering properties of a thermoset network can be combined
with those provided by aggregated ion-pairs.
Fig. 5 provides time-resolved measurements of dynamic storage
modulus (G ) recorded at a fixed temperature, strain amplitude and
b
0
frequency. A premixed BIIR compound containing 1.0 eq BuIm was
tested to gain insight into BIIR þ alkylimidazole reactions under
solvent-free conditions. The rheological data plotted in Fig. 5a show
ꢀ
clear evidence of N-alkylation activity at 100 C, as the storage
1
modulus increased from 96 to 171 kPa over a 60 min period.
H
NMR analysis of the product revealed full conversion of allylic
bromide to imidazolium bromide functionality analogous to
ꢀ
structures 2a,b. As expected, the same reaction carried out at 160 C
0
progressed faster, with G increasing from 62 to 117 kPa in just
10 min (Fig. 5b). Allylic bromide conversion at this point was
complete, yielding an imidazolium content of 0.13 mmol/g, and
0.02 mmol/g of the conjugated diene product of BIIR
dehydrobromination.
ꢀ
Fig. 4. Influence of ion-pair content a. Intrinsic viscosity (A, toluene, 27 C); b. Solid-
state rheology (C G , -tan d, 160 C, 1 Hz, 3 arc).
0
ꢀ ꢀ

J.S. Parent et al. / Polymer 52 (2011) 5410e5418
5415
KOH
N
N
Bu NBr
4
N
NH
N
toluene
Br
+
Br
N
(3)
Br
Br
solvent
free
2
Br
H
N
+
HBr
Scheme 3. BIIR dehydrobromination.
N
Br
N
Br
N
conjugated diene within 10 min, and 43% was converted after
0 min. BIIR contains a small amount of non-nucleophilic acid
(
Bis-imidazolium thermoset)
2
scavengers such as epoxides and calcium stearate to sequester HBr
that is released by the polymer, thereby stabilizing it against acid-
catalyzed degradation. Limited studies of BIIR dehydrobromination
suggest it proceeds by an E1 mechanism involving deprotonation of
allyl cation intermediates [29]. As such, HBr elimination rates are
not affected by base, and should not be accelerated by imidazole-
based nucleophiles.
*
H
Scheme 2. Bis-imidazolium thermoset synthesis.
HBr elimination is an undesirable side-reaction that can
consume significant amounts of allylic bromide at temperatures
above 120 C (Scheme 3). When heated to 160 C in the absence of
BuIm, 19% of the allylic bromide within BIIR was converted to
When compared at equal amounts of imidazole functionality,
bis-imidazole 3 was less reactive than BuIm (Fig. 5). Whereas the N-
ꢀ
ꢀ
alkylation of 1.0 eq BuIm raised the storage modulus 75 kPa at
ꢀ
1
00 C, 0.5 eq of 3 had no effect under identical conditions. This
inactivity at moderate temperatures has practical advantages, since
heat buildup makes it difficult to keep elastomer compounds below
ꢀ
1
00 C during large-scale formulation mixing. On the other hand,
a
ꢀ
cure temperatures on the order of 160 C should, ideally, provide
rapid cross-linking to a high state of cure. The rheology data plotted
ꢀ
in Fig. 5b show that bis-imidazole 3 is sufficiently reactive at 160 C,
yielding a thermoset material after 40 min that was completely
insoluble in organic solvents. The storage modulus of this
compound evolved more slowly than BuIm in the early stages of the
cure, due in part to the lower overall concentration of 3. However,
b
Fig. 5. Storage modulus for BIIR mixtures containing BuIm or bis-imidazole, 3 (a.
Fig. 6. Storage modulus for BIIR mixtures containing 0.5 eq of DABCO, DIPHOS, bis-
ꢀ
ꢀ
ꢀ
0 ꢀ
imidazole 3, and 4,4 -bipyridine (1 Hz, 3 arc).
100 C, b. 160 C; 1 Hz, 3 arc).

5416
J.S. Parent et al. / Polymer 52 (2011) 5410e5418
ꢀ
An ideal thermoset formulation is inactive at 100 C to facilitate
ꢀ
compound mixing, provides a short induction period at 160 C to
H
allow the compound to conform to the dimensions of the
compression mold, and subsequently cross-links quickly toward
a stable final modulus. Of the four formulations examined in Fig. 6,
only the bis-imidazole curative satisfies these three criteria.
DIPHOS was too reactive at 100 C, and did not provide a sufficient
induction period at 160 C. It did, however, provide the same
N
Br (1a)
Br
Br
NH
N
H
N
ꢀ
ꢀ
storage modulus as bis-imidazole, unlike DABCO, which cross-
ꢀ
linked BIIR very quickly at 100 C, and yet evolved toward a rela-
tively low modulus at 160 C. Previous studies of tertiary amine
H
(1b,c)
ꢀ
alkylation by BIIR have shown these reactions to be rapid and
reversible [15]. In context of a DABCO-based formulation, deal-
kylation of the tetraalkylammonium bromide is expected to limit
the ultimate state cure. In the case of bipyridine, inadequate reac-
tivity at 160 C was accompanied by reversion of the storage
modulus with prolonged heating.
base
N
H
ꢀ
Br
N
N
H
Br
N
H
3.4. Multiple N-alkylations of imidazole
A second route to a thermoset imidazolium ionomer involves
the repeated alkylation of imidazole by BIIR in the presence of
a suitable base (Scheme 4). Nucleophilic displacement of bromide
(
4)
Scheme 4. Bis-alkylation of imidazole.
A
from the elastomer yields an imidazolium salt with a pK on the
order of 7. Deprotonation is, therefore, easily accomplished by
moderate bases to give an imidazole functionalized polymer
intermediate whose alkylation by a second allylic halide generates
the desired ionomer cross-link. Our studies of the BIIR þ imidazole
system began with the reaction of imidazole with 2 eq of the model
compound 1, which gave the diallylimidazolium bromide (Scheme
4, 4) as a mixture of three geometric isomers. Although the insol-
ubility of thermosets prohibits NMR analysis of BIIR-derived
material, we suggest that an analogous dialkylation product
underlies an imidazole-based thermoset synthesis.
beyond the 5 min mark of the process, the modulus of the bis-
imidazole formulation exceeded that of BuIm, and approached an
ultimate modulus value twice that of the monofunctional nucleo-
phile. This is evidence of multiple N-alkylations of 3 to give
a thermoset ionomer comprised of covalent cross-links and
a network of aggregated ion-pairs.
Analogues of our bis-imidazole approach are easily conceived,
including formulations based on 1,2-bis(diphenylphosphino)
0
ethane (DIPHOS), 1,4-diazabicyclo[2.2.2]octane (DABCO), and 4,4 -
bipyridine. Quaternization of both nucleophilic centres in these
reagents will yield thermoset products if they are alkylated irre-
versibly, and at a rate that is competitive with BIIR dehy-
drohalogenation. Fig. 6 presents storage modulus data for mixtures
of BIIR containing 0.5 eq of each reagent. These experiments con-
With knowledge of the products of BIIR þ imidazole reactions,
studies shifted to cure dynamics and yields. Our imidazole
formulations included 1,8-bis(dimethylamino)naphthalene, also
Ó
known as Proton Sponge , as an additional acid scavenger. Its
intended role was to sequester HBr released through BIIR dehy-
drohalogenation, thereby minimizing the potential for imidazole
protonation. This non-nucleophilic base had no effect on the
ꢀ
sisted of two phases: a 20 min period at 100 C, followed imme-
ꢀ
diately by 40 min at 160 C.
a
b
ꢀ
Fig. 7. Solid-state storage modulus for BIIR mixtures containing various concentrations of imidazole and Proton Sponge (1 Hz, 3 arc).

J.S. Parent et al. / Polymer 52 (2011) 5410e5418
5417
In every experiment summarized in Fig. 7, the modulus of
compounds comprised of BIIR, imidazole and Proton Sponge was
O
O
O
Ó
O
ꢀ
ꢀ
stableat100 C, and reduced steadily upon heating to160 C forabout
3 min, which marked the onset of cure. Fig. 7a illustrates the effect of
imidazole concentration on cure rates and yields. As expected,
increasing imidazole availability from 0.25 eq to 0.50 eq raised the
cure rate and extent. Going beyond the stoichiometric requirement to
O
+
HO
N
N
N
N
O
(5)
HO
1
.00 eq of imidazole raised the modulus further, likely due to a more
Aliquat 336
KOH
competitive N-alkylation rate relative to BIIR dehydrobromination.
Presumably, higher nucleophile concentrations promote cross-
linking by directing allylic bromide conversion toward dia-
llylimidazolium bromide at the expense of conjugated diene.
As noted above, commercial grades of BIIR are stabilized by non-
nucleophilic acid scavengers and, as such, additional base is not an
absolute requirement for the imidazole reactions of present
interest. Nevertheless, the data plotted in Fig. 7b show that 0.5 eq of
Br
O
toluene
O
O
N
N
IIR-g-Imidazole)
O
(
Ó
Proton Sponge increased the final modulus of a BIIR þ imidazole
H
formulation by 17%. This improvement is attributed to the superi-
Ó
our basicity of Proton Sponge , which serves to limit the extent to
which free imidazole and polymer-bound imidazole are rendered
non-nucleophilic by protonation.
solvent
free
Br
3.5. IIR-g-Imidazole
H
One drawback of the thermoset chemistry described above is
the need to handle hazardous reagents such as imidazole within
conventional polymer compounding devices. An attractive varia-
tion involves the synthesis of a butyl rubber derivative containing
pendant imidazole functionality, which when mixed with BIIR and
heated to standard curing temperatures, provides an imidazolium
thermoset. One example of this approach is illustrated in Scheme 5,
wherein BIIR is reacted with an ambident nucelophile to prepare
IIR-g-imidazole.
O
N
O
O
H
N
Br
O
(IIR-g-Imidazolium thermoset)
Scheme 5. Synthesis and crosslinking of an imidazole-functionalized elastomer.
The challenge in transforming BIIR into a nucleophilic derivative
lies in the potential for polymer-bound imidazole functionality to
react with unconverted allylic bromide, yielding a prematurely
cross-linked product. This led us to use a carboxylate salt of
compound 5 (Scheme 5), which was prepared by reaction of succinic
anhydride and 1-hydroxyethylimidazole. While selectivity for O-
alkylation over N-alkylation was sufficient to avoid large-scale gel
formation, protecting group chemistry, or the activated carbonate
approach recently applied to unhalogenated poly(isobutylene-co-
isoprene) [38], may hold greater promise than the ambident
nucleophile strategy used in this work.
storage modulus of BIIR when mixed into the elastomer and heated
to 160 C. Note that full allylic bromide conversion to bis-
ꢀ
allylimidazolium functionality analogous to 4a,b would require
Ó
0
.5 eq of both imidazole and Proton Sponge , but dehydrohaloge-
nation and the acid scavengers present within BIIR will affect this
stoichiometry.
Nevertheless, IIR-g-Imidazole prepared through Scheme 5 was
sufficiently soluble in toluene to allow for its co-dissolving with an
equal mass of BIIR. Precipitation of the IIR-g-Imidazole þ BIIR
mixture from acetone gave an intimately mixed compound that
ꢀ
was dried under vacuum before cross-linking using the 100 C/
ꢀ
1
60 C procedure described above. The data plotted in Fig. 8 reveal
ꢀ
modest cure activity over 20 min at 100 C, and very little delay
when heated to 160 C. This reactivity is attributed to a direct path
ꢀ
to cross-link formation, since N-alkylation of IIR-g-Imidazole by the
allylic bromide functionality within BIIR yields a covalent cross-link
immediately, unlike the thermoset preparations described above,
where two alkylations are needed to produce the same result.
4. Conclusions
A new class of isobutylene-rich imidazolium bromide ionomers
has been prepared through N-alkylation of a range of imidazole
nucleophiles by the allylic halide functionality within BIIR. BuIm
alkylations proceed irreversibly to give stable ion-pairs whose
aggregation in dilute solution and the solid state affects material
properties in a manner that is consistent with known elastomeric
ꢀ
Fig. 8. Solid-state storage modulus for a 50:50 IIR-g-Imidazole:BIIR mixture (1 Hz, 3
arc).

5418
J.S. Parent et al. / Polymer 52 (2011) 5410e5418
ionomers. The versatility of imidazole chemistry can be exploited to
synthesize thermoset ionomers whose material properties are
derived from a network of covalent cross-links and a network of
ion-pair aggregates. The dynamics of these solid-state reactions are
amenable to standard compounding and cross-linking methods,
yielding unconventional thermosets with dynamic rheological
properties akin to those generated by sulfur and peroxide cure
formulations.
[13] Dugal-Tessier J, Kuhn PS, Dake GR, Gates DP. Heteroatom Chem 2010;21(5):
55e60.
3
[
[
14] Gordon JE. J Org Chem 1965;30:2760e3.
15] Parent JS, White GDF, Whitney RA, Hopkins W. Macromolecules 2002;35:
3374e9.
[16] Chowdhury S, Mohan RS, Scott JL. Tetrahedron 2007;63:2363e89.
[17] Chan BKM, Chang NH, Grimmett MR. Aust J Chem 1977;30:2005e13.
[18] MacKnight WJ, Lundberg RD. Rubber Chem Technol 1984;57:652e63.
[19] Antony P, De SK. J Macromol Sci. Polym Rev 2001;C41:41e77.
[
20] Weiss RA, MacKnight WJ, Lundberg RD, Mauritz KA, Thies C, Brant DA. ACS
symposium series. Coulombic Interact Macromol Syst 1986;302:2e19.
21] Parent JS, Thom DJ, White G, Whitney RA, Hopkins WJ. Polym Sci. Part A
Polym Chem 2001;29:2019e26.
[
Acknowledgments
[
[
[
22] So YH. Macromolecules 1992;25:516e20.
23] Huggins ML. J Am Chem Soc 1942;64:2716e8.
24] Chu CY, Vukov R. Macromolecules 1985;18:1423e30.
Support from the Ontario Centres of Excellence is greatly
appreciated.
[25] Chu CY, Watson KN, Vukov R. Rubber Chem Technol 1987;60:636e46.
[
[
26] Yuan J, Antoniettia M. Polymer 2011;52(7):1469e82.
27] Guillen-Castellanos SA, Parent JS, Whitney RA. Macromolecules 2006;39:
2514e20.
References
[28] (a) Parent JS, Malmberg S, McLean JK, Whitney RA. Eur Polym J 2010;46:
[
[
[
[
1] Halliday L. Ionic polymers. 1st ed. New York: Wiley; 1975.
702e8;
2] Eisenberg A, Kim JS. Introduction to ionomers. 1st ed. New York: Wiley; 1998.
3] MacKnight WJ, Earnest TR. J Polym Sci. Macromol Rev 1981;16:41e122.
4] Fitzgerald JJ, Weiss RA. J Macromol Sci. Rev Macromol Chem Phys 1988;C28:
(b) Parent JS, Malmberg SM, Whitney RA. Green Chem 2011;13:2818e24;
(c) Whitney RA, Penciu A, Parent JS, Resendes R, Hopkins W. Macromolecules
2005;38:4625e9.
9
9e185.
5] Cheng S, Beyer FL, Mather BD, Moore RB, Long TE. Macromolecules dx.doi.org/
0.1021/ma200829h.
[29] McLean JK, Guillén-Castellanos SA, Parent JS, Whitney RA, Kulbaba K,
Osman A. Ind Eng Chem Res 2009;48:10759e64.
[30] Malmberg S, Parent JS, Pratt DA, McLean JK, Whitney RA. Macromolecules
2010;43:8456e61.
[31] Lundberg RD, Makowski HS. J Polym Sci. Polym Phys Ed 1980;18:1821e36.
[32] Lundberg RD, Phillips RR. J Polym Sci. Polym Phys Ed 1982;20:1143e54.
[33] Tant MR, Wilkes GL, Kennedy JP. J Appl Polym Sci 1989;37:2873e95.
[34] Capek I. Adv Colloid Interface Sci 2004;112:1e29.
[
1
[
[
[
6] Charnley M, Textor M, Acikgoz C. React Funct Polym 2011;71(3):329e34.
7] White JD, Wilker JJ. Macromolecules 2011;44:5085e8.
8] Thomas RM, Lightbown IE, Sparks WJ, Frolich PK, Murphee EV. Ind Eng Chem
1940;32:1283e92.
[
9] Arjunan P, Wang HC. Polym Mater Sci Eng 1997;76:310e1.
[
10] Parent JS, Penciu A, Guillen-Castellanos SA, Liskova A, Whitney RA. Macro-
[35] Xie HQ, Xu Die JG. Angew Makromol Chem 1990;114:177e87.
[36] Coran AY. Vulcanization. In: Mark JE, Erman B, Eirich FR, editors. Science and
technology of rubber. San Diego: Academic Press; 1994. p. 339e86.
[37] Bagrodia S, Tantt MR, Wilkes GL, Kennedy JP. Polymer 1987;28:2207e26.
[38] Bonduelle CV, Gillies ER. Macromolecules 2010;43:9230e3.
molecules 2004;37:7477e83.
[11] Tsou AH, Duvdevani I, Agarwal PK. Polymer 2004;45(10):3163e73.
12] Parent JS, Liskova A, Whitney RA, Resendes RJ. Polym Sci. Part A Polym Chem
[
2005;43(22):5671e9.