Evaluation of different catalyst systems for bulk polymerization through "click" chemistry

Article abstract of DOI:10.1016/j.polymer.2011.07.018

"Click" chemistry is a frequently used technique in polymer chemistry for coupling polymer end groups to construct novel copolymer architectures and as a step-growth polymerization technique. In this study, the bulk polymerization of a diyne and a bisazide were achieved through the copper-catalyzed alkyne-azide 1,3-cycloaddition (CuAAC) with different catalysts, including the homogeneous catalyst Cu(PPh₃)₃Br and a heterogeneous Cu/C catalyst. The effects of different catalyst systems on the kinetics of the "click" polymerization were evaluated by differential scanning calorimetry (DSC) and oscillatory rheology. Additionally, the thermomechanical properties of resulting polymers were evaluated by temperature-ramp oscillatory rheology and thermal stabilities by thermogravimetry.

Full text of DOI:10.1016/j.polymer.2011.07.018

Polymer 52 (2011) 4435e4441
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Evaluation of different catalyst systems for bulk polymerization through
“click” chemistry
*
Xia Sheng, Darman M. Rock, Timothy C. Mauldin, Michael R. Kessler
Materials Science and Engineering Department, Iowa State University, Ames, IA 50011, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
“Click” chemistry is a frequently used technique in polymer chemistry for coupling polymer end groups
to construct novel copolymer architectures and as a step-growth polymerization technique. In this study,
the bulk polymerization of a diyne and a bisazide were achieved through the copper-catalyzed alkyne-
azide 1,3-cycloaddition (CuAAC) with different catalysts, including the homogeneous catalyst
Cu(PPh3)3Br and a heterogeneous Cu/C catalyst. The effects of different catalyst systems on the kinetics of
the “click” polymerization were evaluated by differential scanning calorimetry (DSC) and oscillatory
rheology. Additionally, the thermomechanical properties of resulting polymers were evaluated by
temperature-ramp oscillatory rheology and thermal stabilities by thermogravimetry.
Received 23 March 2011
Received in revised form
13 July 2011
Accepted 16 July 2011
Available online 23 July 2011
Keywords:
“Click” chemistry
Kinetics
Ó 2011 Elsevier Ltd. All rights reserved.
Rheology
1. Introduction
have reported tens of different linear and cross-linked “click”
polymers, prepared in both bulk and solution conditions. Other
The alkyneeazide cycloaddition was first introduced as an
uncatalyzed reaction by Huisgen et al. [1] and is often classified as
a “click” chemistry, a termed coined by Sharpless and co-workers
novel polymers, such as conjugated polymers [27] and
perfluorocyclobutyl-containing polymers [28], were also synthe-
sized efficiently through the CuAAC.
[2], when catalyzed with
a
copper (I) complex. Sharpless
Any copper molecule in the þ1 oxidation state will catalyze the
“click” reaction, but different ligands on the catalyst do influence
reaction kinetics [29]. The Cu(PPh3)3Br, made of Cu(I)Br and tri-
phenylphosphine ligand, is a commonly used catalyst in the CuAAC
due to its solubility in organic monomers, high catalytic efficiency
and low cost [30,31], Lipshutz et al. [32] described an efficient,
copper-in-charcoal (Cu/C) heterogeneous catalyst for the “click”
reaction. With the addition of triethylamine (Et3N) to this catalyst,
alkyne-azide cycloadditions yielded a library of 1,4-disubstituted-
triazole products in very high yields (w99%) and functional group
tolerance [33]. This especially high-yielding variation of the CuAAC
makes the Cu/C catalyst a prime candidate for use in a “click”
polymerization, of which the step-growth nature of this reaction
makes high yields a necessity to generate high molecular weight
polymer. In addition, the residual charcoal in which the Cu(I)
species was immobilized can also serve as a fundamentally inter-
esting “multifunctional catalyst”, in which it may act as a reinforc-
ing particle to fabricate an in-situ generated polymer-matrix
composite.
proposed a set of criteria to define “click” chemistryda chemical
reaction that is efficient, robust and can be conducted under simple
conditions (e.g. air/moisture insensitive, solvent-free, etc.) while
generating only inoffensive byproducts that can be easily removed
(i.e. minimal chromatographic work-up). The copper-catalyzed
alkyne-azide 1,3-cycloaddition (CuAAC), which was reported
independently by the groups of Sharpless [3] and Meldal [4], is
a typical “click” reaction (Fig. 1) and has become synonymous with
“click” chemistry because of its fulfillment of the above criteria.
The CuAAC is a powerful technique in polymer chemistry and
materials science. Many macromolecular structures were success-
fully constructed by the CuAAC, including block copolymers [5e8],
macro-cyclic rings [9,10], graft brush and block copolymers [11e15],
stars [16e18], and dendrimers [19e21]. “Click” chemistry has also
been used successfully for surface and interface modification
[22e24]. In addition, “click” chemistry has been used periodically
as a step-growth polymerization mechanism. Finn et al. [25,26],
In this study, bisphenol A-based bisazide (BA-bisazide) and
bisphenol E-based diyne (BE-diyne) monomers were synthesized
and bulk “click” polymerizations were conducted (Fig. 2). To
compare the effect of different catalytic systems, a BE-diyne/BA-
* Corresponding author.
E-mail address: mkessler@iastate.edu (M.R. Kessler).
0032-3861/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2011.07.018

4436
X. Sheng et al. / Polymer 52 (2011) 4435e4441
with BA-bisazide in a 1:1 M ratio at room temperature using
magnetic stirring. For the catalyzed samples (samples 2, 3, 4 and 5;
also described in Table 1), the molar ratio of BE-diyne to BA-
bisazide remained stoichiometric, and the Cu/C and Cu(PPh3)3Br
catalysts were added in a 1:20 and a 1:160 M ratio (with respect to
diyne), respectively. For samples 3 and 5, Et3N was added at
a 1:20 M ratio with respect to diyne. All samples were magnetically
stirred to ensure satisfactory mixing of components. Samples
containing Cu/C catalyst were then submerged in an ultrasonic bath
to disperse the catalyst. Prepared samples were then moved to
a freezer at ꢀ13 ꢁC in sealed vials for subsequent characterizations.
Dynamic cure experiments were performed using differential
scanning calorimetry (DSC) on a model Q20 DSC (TA Instruments,
Inc., New Castle, Delaware). The DSC cell was purged with nitrogen
gas at a flow rate of 50 mL/min for all DSC experiments. Uncured
samples were loaded into hermetic aluminum DSC pans using
apipette thensealed witha lid. A smallholewasmadeinthecenterof
each lid to prevent pressure buildup during heating. Uncured
samples were tested from ꢀ50 to 250 ꢁC in the dynamic scan mode at
heating rates of 2, 5, 10, and 15 ꢁC/min. This temperature range
captured the complete exothermal peaks, indicating full conversion,
and was below the decomposition temperature of all samples. After
cooling to room temperature in the DSC cell, another dynamic scan
was immediately performed at a rate of 10 ꢁC/min from ꢀ50 to 150 ꢁC
in order to observe the glass transition temperature of each sample.
An AR2000ex stress-controlled rheometer (TA Instruments,
Inc.) was used to monitor the rheological properties of uncured
samples, as well as the thermomechanical properties of resulting
polymers. A parallel plate geometry (25-mm diameter disposable
aluminum) was used with a gap of 0.5 mm. An uncured sample
(0.25 mL) was transferred to the parallel plates using a 1-mL
syringe, and then tested at an isotherm of 120 ꢁC in oscillatory
time-sweep mode at amplitude of 1000 Pa and 1 Hz. After holding
at 120 ꢁC for another hour, an oscillatory temperature-ramp step
was followed: the temperature was ramped down from 120 to
25 ꢁC at cooling rate of 1 ꢁC/min from which the thermomechanical
properties of resulting polymer were evaluated.
Fig. 1. The Cu(I)-catalyzed azide alkyne cycloaddition (CuAAC) reaction.
bisazide sample without catalyst was prepared and tested as
a benchmark. Two catalyst systems were evaluated in this study:
Cu(PPh3)3Br and Cu/C catalyst. Triethylamine (Et3N) was used to
accelerate the “click” reactions with both the homogeneous and
heterogeneous catalyst systems. The catalytic ability of each cata-
lyst system was evaluated through the analysis of “click” bulk
polymerization kinetics via differential scanning calorimetry (DSC)
and oscillatory rheology. The thermomechanical properties of the
resulting polymers were investigated with oscillatory rheology as
well. The thermal stabilities of the resulting polymers made with
different catalysts were tested with thermogravimetry.
2. Experimental
2.1. Materials
Diglycidyl ether of bisphenol A (DGEBA) epoxy resin was
purchased from Miller-Stephenson Chemical Co. (Danhury, CT) as
EPONÔ resin 828. All other chemicals were purchased from Sig-
maeAldrich (St. Louis, MO) and used as-received, unless otherwise
stated. The synthesis procedures and characterizations of bisphenol
A bisazide (BA-bisazide) and bisphenol E diyne (BE-diyne) are
described elsewhere in our previous publication [34]. The synthetic
routes to the BA-bisazide and BE-diyne are shown in Fig. 2. The Cu/
C catalyst was prepared by following the reported procedure
without modification [32].
2.2. Sample preparation and characterization
Five different catalyst formulations were used in this study. The
uncatalyzed sample (sample 1) was prepared by mixing BE-diyne
Fig. 2. Synthesis routes to monomers and “click” bulk polymerization.

X. Sheng et al. / Polymer 52 (2011) 4435e4441
4437
Table 1
Summary of the test results of the BE-alkyne/BA-azide sample with different catalysts.
Sample
Catalyst
Tonset (ꢁC)
DH
b (J/g)
tonset (s)
a
a
Tp (ꢁC)
b
Tg (ꢁC)
c
Td (ꢁC)
1
2
3
4
5
w/o
97.6
67.1
57.5
97.3
86.8
139.9
95.7
95.9
139.5
121.6
98.8 ꢂ 2.4
85.9 ꢂ 1.8
95.7 ꢂ 2.7
97.0 ꢂ 2.7
99.5 ꢂ 2.6
621.7 ꢂ 5.5
584.5 ꢂ 16.3
471.9 ꢂ 22.8
554.7 ꢂ 3.3
566.8 ꢂ 5.7
338.5
318.0
332.5
338.9
338.0
1310
90
Cu(PPh3)3Br
Cu(PPh3)3Br þ Et3N
Cu/C
Cu/C þ Et3N
ed
910
250
a
Taken from dynamic scans at 10 ꢁC/min via DSC.
Average values and standard deviations from the samples polymerized at different heating rates.
At 5% mass loss (TGA).
b
c
d
Viscosity began increasing right after data collection was started and the accurate value could not be determined (Rheology).
The polymers were obtained as transparent, yellow-orange
reaction heat,
(with respect to time). The values of
different heating rates are slightly different. The reaction heats
from samples with different catalyst systems are summarized in
Table 1.
D
H, can be calculated by the integration of the peak
H from the dynamic scan at
solids. Nuclear magnetic resonance (NMR) spectra of the poly-
merized samples were obtained on a 300 MHz spectrometer with
residual protonated solvent used as an internal reference. 1H NMR
D
(chloroform-d), d: 1.58 (br m, eCH3), 3.92 (br m, eCH2eOH), 4.06
(br m, >CHeCH3), 4.64 (br m, >CHeCH2eOAr), 5.02 (br m,
>CHetriazole), 5.13 (br m, ArOeCH2etriazole), 6.65e6.91 (br m,
AreH), 7.11 (br m, AreH), 7.71 (br m, triazoleeH).
Gel permeation chromatography (GPC) was performed on
a Viscotek GPC Max 280 with both a refractive index and UV
detector (256 nm). Analyses were performed using THF eluent at
a flow rate of 1.0 mL/min. In all cases, both detectors yielded
satisfactorily similar molecular weight information, and data from
the refractive index detector is reported. Calibration was based on
polystyrene standards. The molecular weight distribution from GPC
Without addition of any catalyst to the BE-diyne/BA-bisazide
system, onset temperatures are not significantly affected by the
applied heating rates, indicating that the uncatalyzed “click”
polymerization can be triggered at around 100 ꢁC. To clarify the
effects of different catalysts, the DSC thermograms at 10 ꢁC/min
from all samples are plotted together in Fig. 4. As expected, the
uncatalyzed reaction (sample 1) gives the highest Tonset value
(97.6 ꢁC), and addition of catalysts lowers Tonset. The Cu(PPh3)3Br
catalyst, samples 2 and 3, show a significant reduction in Tonset: 30
and 40 ꢁC, respectively. The catalytic effect of Cu/C system is not as
pronounced as Cu(PPh3)3Br; the Tonset of sample 4 is nearly iden-
tical to that of sample 1. Triethylamine shows noticeable effects on
the onset temperature; for both the Cu(PPh3)3Br and Cu/C catalyst
systems, addition of Et3N decreases Tonset by about 10 ꢁC. In addi-
tion to these changes in onset temperatures, the peak temperatures
and shapes are also affected by the different catalyst systems. The
thermogram of sample 2 (Cu(PPh3)3Br) appears to be a sum of two
separate peaks, indicating competing catalyzed and uncatalyzed,
thermal triggered “click” reaction mechanisms. It should be noted
that while our “click” polymerizations are slower than expected
based on the “click” chemistry methodology, bulk “click” poly-
merizations sometimes tend to have slightly slower reaction
kinetics [25,26,36].
is shown in the Supplementary information: Mn
Mw ¼ 46000, PDI ¼ 1.89.
¼
26000,
FTIR was performed on a Bruker IFS 66 v/s spectrometer. Two
milligrams of polymer sample were ground with 100 mg of KBr into
a fine powder and pressed into a pellet for the transmission
measurement. The mid-infrared absorption spectra of the powder
were recorded in the range of 4000e400 cmꢀ1, obtained using 32
scans at
a 4
cmꢀ1 resolution. Spectra (available in the
Supplementary information) revealed the absence of the azide and
alkyne functional groups present on the monomers.
The thermal decomposition of the cured samples was investi-
gated using thermogravimetric analysis (TGA) on a model Q50 TGA
(TA Instruments, Inc.). Cured samples were obtained from the filled
aluminum pans after DSC experiments were complete. About 6 mg
of cured sample was placed in a platinum pan and heated from
room temperature to 800 ꢁC at a rate of 20 ꢁC/min with an air purge
rate of 60 mL/min.
Typically, peak temperatures shift to higher temperatures with
increasing heating rates. Kissinger’s equation allows various kinetic
146.6 °C•
3.0
2.5
2.0
1.5
1.0
0.5
0.0
2
5
°
°
10
C/min
C/min
3. Results and discussion
°
°
C/min
C/min
Performing dynamic scanning at multiple heating rates is
a convenient and useful method to investigate the kinetics of
chemical reactions, particularly for bulk polymerizations. From
a series of multiple heating rate dynamic scans, many parameters
15
139.9 °C•
can be obtained directly, such as the reaction heat (DH), onset
temperatures (Tonset) and peak temperatures (Tp) of the exothermic
reaction peak, and the glass transition temperatures (Tg) of result-
ing polymers. Further analysis of the experiment data could yield
useful parameters such as activation energy (E), variances in acti-
vation energy with conversion, reaction orders, etc. Therefore, the
kinetics of the “click” bulk polymerization is mainly investigated
herein by conducting a series of dynamic scans via DSC.
128.5
114.0
°C•
°C•
-50
0
50
100
150
C)
200
250
BE-diyne/BA-bisazide samples (without or with catalyst) were
heated in DSC cell from ꢀ50 to 250 ꢁC at heating rates of 2, 5, 10 and
15 ꢁC/min. Fig. 3 shows a series of typical dynamic scan thermo-
grams for sample 1 (without catalyst [35]). The exothermic peaks
represent the heat generated during the “click” reaction. The total
Temperature (
°
Fig. 3. Typical DSC dynamic scan thermograms at different heating rates for sample 1
(BE-alkyne/BA-azide without catalyst).

4438
X. Sheng et al. / Polymer 52 (2011) 4435e4441
Table 2
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
Summary of the activation energies and other kinetic parameters.
w/o cat.
Sample
Catalyst
Ea (kJ/mol)
ln Ab
nb
mb
n þ mb
Cu(PPh3)3Br
1
2
3
4
5
w/o
76.9
63.0
61.5
76.7
66.3
17.6
15.2
15.5
17.6
15.2
1.4
1.8
1.8
1.4
1.7
0
1.4
1.9
2.0
1.4
1.8
Cu(PPh3)3Br+Et3N
Cu/C
Cu/C+Et3N
Cu(PPh3)3Br
Cu(PPh3)3Br þ Et3N
Cu/C
Cu/C þ Et3N
0.1
0.2
0
0.1
a
Calculated from Kissinger’s analysis.
Average values from the fitting parameters of original experimental data.
b
bring reactive groups to the charcoal surfaces, but such molecular
mobility certainly diminishes with reaction progress in a bulk
polymerization. However, the catalytic effect is more pronounced
upon addition of triethylamine, evidenced by the decrease in acti-
vation energy in sample 5 (Cu/C þ Et3N). In general, the observed
trend between activation energy with the different catalyst systems
is consistent with the trend of onset temperature changes.
The DSC evaluation not only gives the values of activation
energy but also indicates further details of reaction kinetics
through model fitting. The basic kinetic rate equation can be
expressed as:
-50
0
50
100
150
C)
200
250
Temperature (
°
Fig. 4. DSC thermograms at 10 ꢁC/min for the BE-diyne/BA-bisazide system containing
different catalysts.
parameters (e.g. activation energy (E)) to be extracted from these
peak temperature shifts [37]:
d
a
¼ kðTÞf ð Þ
a
(2)
) is
!
ꢀ
ꢁ
dt
b
Tp2
AR
E
E
RTp
ln
¼ ln
ꢀ
(1)
where
a is the conversion, T is the absolute temperature in K, f(a
a function related to the reaction type and conversion, and k(T) is
where
b
is the heating rate, R is the universal gas constant, and A is
the rate constant described by the Arrhenius equation:
the pre-exponential factor. A plotof lnð
b
=Tp2Þ vs.1/Tp yields values for
ꢀ
ꢁ
Ea
RT
the pre-exponential factor A and activation energy E from the
intercept and the slope of the linear fit, respectively. Such a plot is
shown in Fig. 5 for sample 1, from which the activation energy is
estimated as 76.6 kJ/moldhigher than other samples with catalysts
(Table 2). As expected from the lower onset temperatures,
Cu(PPh3)3Br did indeed catalyze the “click” bulk polymerization,
indicated by the relatively lower activation energies of samples 2
and 3. The effect of the triethylamine, however, is not as pronounced
as would be expected by the decrease in onset temperature between
samples 2 and 3, as the E of sample 3 is just slightly lower than
sample 2. For the Cu/C catalyst system sample 4 does not signifi-
cantly impact the activation energy, relative to the sample without
any catalyst (sample 1). It is possible that the lack of a significant
catalytic effect in sample 4 is due to the fact that Cu/C is a hetero-
geneous catalyst, hence requiring significant molecular mobility to
kðTÞ ¼ Aexp ꢀ
(3)
Ea is the activation energy at a given conversion.
Two common reaction modelsdautocatalytic (equation (4)) and
n-th order (equation (5))dwere used to further analyze the
experimental data:
f ð
a
a
Þ ¼ ð1 ꢀ
Þ ¼ ð1 ꢀ
a
Þ
Þ
nam
(4)
n
f ð
a
(5)
where, n and m are reaction orders. If m equals to zero, the auto-
catalytic model collapses into the n-th order model. Fitting the data
to these reaction models can be achieved by monitoring heat
evolution during the reaction, which can be directly measured by
DSC and calculated as below:
-13
dH
dt
d
a
d t
¼
D
H$
(6)
By combining with equations (2)e(4), equation (6) is expressed as:
ꢀ
ꢁ
-14
dH
dt
Ea
RT
¼
D
HAexp ꢀ
ð1 ꢀ
a
Þ
nam
(7)
To simplify the calculation, the constant value of activation energy
(E) from Kissinger’s analysis is taken as an approximation to replace
the Ea in equation (7). The pre-exponential factor A and reaction
orders (n and m) are determined by the best fitting results. Fig. 6
shows the autocatalytic model fits of sample 1, where the experi-
mental heat flows are normalized with respect to heating rate. In
most cases, the model curves fit the experimental data well
throughout the entire temperature range. Although the dynamic
scan thermograms of the sample 2 apparently have two peaks (or
one major peak with shoulder), the model fit is mostly consistent
with the experimental data with a coefficient of determination
R2 > 0.95.
β
-15
Eα=76.6 kJ/mol
-16
0.0023
0.0024
0.0025
1/T (1/K)
0.0026
p
Fig. 5. Kissinger’s analysis for sample 1.

X. Sheng et al. / Polymer 52 (2011) 4435e4441
4439
The kinetic parameters of all catalyst systems are summarized in
40
35
30
25
20
15
10
5
1000
800
600
400
200
0
Table 2. For sample 1, the reaction order m is zero, and the reaction
follows an n-th order kinetics with a reaction order n of 1.4. Sample
4 yields similar kinetic parameters with sample 1, indicating that
Cu/C (without the addition of triethylamine) is non-catalytic, likely
for the reasons mentioned earlier. The addition of triethylamine
activates the catalytic effect of the Cu/C catalyst systemdthe
kinetic parameters from sample 5 have similar values as sample 2.
The Cu(PPh3)3Br catalyst shows a larger reaction order (n ¼ 1.8 for
both samples 2 and 3), which lends evidence to its catalytic effect
on the “click” bulk polymerization of BE-diyne/BA-bisazide. The
reaction order m is a nonzero number of 0.1 and 0.2 for samples 2
and 3, respectively, indicating a complex reaction mechanism.
The glass transition temperatures of the resulting polymers
were determined via DSC at a heating rate of 10 ꢁC/min immedi-
ately after each bulk polymerization (Table 1). With the exception
of sample 2 (Tg ¼ 85.9 ꢁC), all of the catalyst systems yielded
G'
G"
*|
|η
•
η
0
0
1000
2000
3000
4000
Time (s)
polymers
with
similar
glass
transition
temperatures
Fig. 7. Typical time-sweep oscillatory rheology experiment (sample 1 at 120 ꢁC for
(Tg ¼ 95.7e98.8 ꢁC).
1 h).
In addition to monitoring the reaction via heat evolution in the
DSC, dynamic oscillatory rheology was used to observe changes in
rheological properties during isothermal cure. The thermo-
mechanical properties of the resulting polymers were investigated
with rheology as well. To evaluate the rheological properties of the
samples containing different catalysts, a two-step procedure was
applied: a time-sweep and a temperature-ramp step. First, the
time-sweep oscillatory experiment was performed. The prepared
monomers/catalyst mixture was isothermally cured at 120 ꢁC for
1 h under an oscillatory shear stress of 1000 Pa, and the evolution of
rheological properties were recorded in real time. Second, the
temperature-ramp oscillatory experiment immediately followed
the isothermal curing. The temperature was ramped down at
a cooling rate of 1 ꢁC/min from 120 to 25 ꢁC in order to evaluate the
thermomechanical properties of the resulting polymers.
In a time-sweep curing test at isothermal temperature, storage
shear modulus (G0), loss shear modulus (G000) and dynamic viscosity
(jh*j) generally increase with the time. G is usually much lower
than G00 at the early stages of isothermal curing, but it increases at
a faster rate than does G00. At the gel time, G0 is equal to G00 (G0 ꢃ G00
crossover), indicating the formation of a cross-linked network
[38aec]. However, the “click” polymerization of BE-diyne/BA-
bisazide results in a linear polymer with a glass transition
temperature at maximum conversion (100 ꢁC) below the
isothermal cure temperature of 120 ꢁC. Therefore, even after
complete polymerization (1 h), the resulting polymer is in its
rubbery state and the value of G0 is expected to be lower than G00
(Fig. 7). For our purposes, the viscosity increase is deemed a good
indication of the “click” polymerization progress.
The viscosity evolution during isothermal cure is plotted in
Fig. 8. If the onset of viscosity increasing is taken as the start of the
“click” polymerization, the time (tonset) at such a point is an indi-
cation of the catalytic activity of each catalyst. While the “click”
polymerization surely begins by 120 ꢁC, the viscosity of sample
increases at a relatively slow rate because step-growth polymeri-
zation yields the high molecular weight polymerdtherefore,
higher viscositydonly at higher conversions. For example, the
viscosity of the system without catalyst begins to show measurable
increases in viscosity only after about 20 min (Table 1). The relative
catalytic effect of the Cu/C catalyst is much more pronounced here
(sample 4) than in the DSC results, indicated by the increase of
viscosity at around 15 min and a significantly higher viscosity value
at 1 h than the system without any catalyst. However, sample 5,
which contains Cu/C and Et3N, does not show significant differ-
ences from sample 1. It is possible that the addition of Cu/C yields
a rheopectic solutionddefined as a solution exhibiting an increase
in viscosity over time under a constant shear force. The Sample 2
(with Cu(PPh3)3Br catalyst) shows a rapid increase in viscosity
14
12
10
8
2
5
10
15
°
°
C/min
C/min
w/o cat.
14
12
10
8
°
C/min
C/min
Cu(PPh3)3Br
°
Cu(PPh3)3Br+Et3N
Cu/C
Cu/C+Et3N
increase of
heating rate
6
6
4
4
2
Cu/C+Et3N
2
0
0
300
350
400
450
500
w/o cat.
Temperature (K)
0
500 1000 1500 2000 2500 3000 3500 4000
Time (s)
Fig. 6. The autocatalytic model fits of sample 1 at different heating rates, where the
heat flow is normalized with respect to the heating rate. Solid lines: autocatalytic
model fits; Symbols: experimental data.
Fig. 8. Time-sweep test results of rheology oscillatory experiments at 120 ꢁC for 1 h.

4440
X. Sheng et al. / Polymer 52 (2011) 4435e4441
10
1
10
w/o cat.
100
G'
Cu(PPh3)3Br
Cu(PPh3)3Br+Et3N
Cu/C
Cu/C+Et3N
80
1
G"
0.1
60
δ
98
0.01
1E-3
1E-4
40
0.1
96
94
G'
tan
δ
20
G"
tan
92
δ
90
0.01
300
0
320
340
360
20
40
60
80
100
120
Temperature (°C)
100 200 300 400 500 600 700 800
Temperature ( C)
Fig. 9. Typical oscillatory temperature-ramp test results for sample 1 from 120 to 25 ꢁ
C
°
at 1 ꢁC/min.
Fig. 11. Thermogravimetric analysis of both catalyzed and uncatalyzed polymers.
early, followed by a relatively slower increase at longer times.
Sample 3 (with Cu(PPh3)3Br and Et3N) shows a rapid increase in
viscosity immediately after data collection is started and results in
the highest viscosity after an isothermal curing time of 1 h.
After the isothermal time-sweep experiment was finished, the
temperature was ramped down at a cooling rate of 1 ꢁC/min from
120 to 25 ꢁC. Typical temperature-ramp test results are shown in
Fig. 9 for the polymerized sample 1. The Tg can be determined by
each system.
A
lower or broader G00 peak suggests greater
constraints are present in the amorphous phase and longer poly-
mer chains that cause more entanglements [39]. For samples 4 and
5, the situation is more complex. Sample 5, with Cu/C þ Et3N, has
a lower and broader peak, indicating the possibility of a polymer
with higher molecular weight, however, its Tg is lower than sample
4. Potentially contributing to this counterintuitive data is the fact
that the charcoal powder of the Cu/C catalyst also acts as composite
filler that likely affects its rheological properties; this complex
phenomenon will be investigated further in the future.
the onset of the G0, the peak of G00, or the tan
d peak. As mentioned
earlier, at elevated temperatures the linear “click” polymer exists as
a polymer melt at the start of the temperature-ramp, and G00
(viscous component) is much higher than G0 (elastic component),
The effects of the different catalysts on the thermal stability of
the polymers were examined using thermogravimetry (TGA). In the
TGA analysis, the samples were heated from room temperature to
800 ꢁC at a heating rate of 20 ꢁC/min under an air purge. Fig. 11
shows the TGA curves for all samples. The temperature at 5%
mass loss is taken as the decomposition temperature (Td) and the
results are summarized in Table 1. Overall, all samples follow
a similar trend in mass loss, which contains two major degradation
steps. Samples 1, 4, and 5 have identical Td values of 338 ꢁC, and the
triethylamine does not significantly affect the decomposition of the
system with Cu/C catalyst. The samples with Cu(PPh3)3Br catalyst (2
and 3) have a lower Td. Interestingly, the triethylamine does affect
the decomposition temperature of the system with the Cu(PPh3)3Br
catalyst. Compared to the control sample 1, Td decreases 20 ꢁC for
the sample 2 without Et3N, and only 6 ꢁC for the sample 3 with
Et3N.
resulting in a large value of tan d. At decreasing temperatures, the
polymer transitions from the rubbery state to the glassy state, and
G0 becomes larger than G00.
Fig. 10 shows the variation of loss shear modulus (G00) with
temperature for polymers made with the different catalyst systems.
Interestingly, the Tg of sample 1 is around 88 ꢁC, which, with the
exception of sample 5, is lower than the Tg’s of catalyzed systems
(around 97 ꢁC). Data derived from G00 is more sensitive to molecular
motions in the polymer chain than what can be derived from G0. For
example, even though the peak temperatures are similar for cata-
lyzed samples 2, 3 and 4, the peak height and width are different for
w/o cat.
1000
800
600
400
200
0
Cu(PPh3)3Br
4. Conclusions
Cu(PPh3)3Br+Et3N
Cu/C
Cu/C+Et3N
The bulk polymerization of BE-diyne/BA-bisazide through
“click” chemistry was successfully conducted using different cata-
lyst systems. Kinetic evaluation showed that the “click” polymeri-
zation could be triggered at elevated temperature without addition
of catalyst. Overall, the catalytic effect of the Cu(PPh3)3Br is much
more pronounced than the heterogeneous Cu/C catalyst. Upon
addition of triethylamine to samples containing the Cu(PPh3)3Br
catalyst, the reaction onset temperature decreased with the addi-
tion of triethylamine while the peak temperature remained
constant. For samples containing Cu/C catalyst, triethylamine
affects the polymerization. Reaction model fitting suggests that the
“click” polymerization follows n-th order reaction kinetics or
autocatalytic kinetics with a very small value of the reaction order
m. The rapid increase of viscosity under isothermal conditions also
40
60
80
100
120
Temperature (°C)
Fig. 10. Loss shear modulus of the polymers made with different catalysts.

X. Sheng et al. / Polymer 52 (2011) 4435e4441
4441
[13] Quémener D, Hellaye ML, Bissett C, Davis TP, Barner-Kowollik C, Stenzel MH.
J Polym Sci Polym Chem 2008;46:155e73.
[14] Jiang X, Vogel EB, Smith MR, Baker GL. Macromolecules 2008;41(6):1937e44.
[15] Tsarevsky NV, Bencherif SA, Matyjaszewski K. Macromolecules 2007;40(13):
4439e45.
lends support to the catalytic effect of the Cu(PPh3)3Br catalyst
toward “click” polymerization. The different catalyst systems do not
have a significant effect on the thermal decomposition behavior of
the resulting polymers with decomposition temperatures ranging
from 318 to 330 ꢁC.
[16] Altintas O, Yankul B, Hizal G, Tunca U. J Polym Sci Polym Chem 2006;44(21):
6458e65.
[17] Gao H, Matyjaszewski K. Macromolecules 2006;39(15):4960e5.
[18] Altintas O, Hizal G, Tunca U. J Polym Sci Polym Chem 2006;44(19):5699e707.
[19] Mynar JL, Choi T-L, Yoshida M, Kim V, Hawker CJ, Fréchet JMJ. Chem Commun;
2005:5169e71.
[20] Goodwin AP, Lam SS, Fréchet JMJ. J Am Chem Soc 2007;129(22):6994e5.
[21] Almutairi A, Guillaudeu SJ, Berezin MY, Achilefu S, Fréchet JMJ. J Am Chem Soc
2008;130:444e5.
[22] Opsteen JA, Brinkhuis RP, Teeuwen RLM, Löwik DWPM, van Hest JCM. Chem
Commun; 2007:3136e8.
[23] Sun XL, Stabler CL, Cazalis CS, Chaikof EL. Bioconjug Chem 2006;17(1):52e7.
[24] Binder WH, Sachsenhofer R. Macromol Rapid Commun 2008;29(12e13):
952e81.
Acknowledgments
Financial support from the US Army Corps of Engineers, Engi-
neer Research and Development Center Construction Engineering
Research Laboratory is gratefully acknowledged. The work was
conducted under contract number W9132T-09-C-0003. The COTR
was Mr. Jonathan Trovillion.
Appendix. Supplementary material
[25] Díaz DD, Punna S, Holzer P, McPherson AK, Sharpless KB, Fokin VV, et al.
J Polym Sci Polym Chem 2004;42(17):4392e403.
[26] Liu Y, Díaz DD, Accurso AA, Sharpless KB, Fokin VV, Finn MG. J Polym Sci
Polym Chem 2007;45(22):5182e9.
[27] van Steenis DJVC, David ORP, van Strijdonck GPF, van Maarseveen JH,
Reek JNH. Chem Commun; 2005:4333e5.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.polymer.2011.07.018.
[28] Zhu Y, Huang Y, Meng W, Lin H, Qing F. Polymer 2006;47(18):6272e9.
[29] Himo F, Lovell T, Hilgraf R, Rostovtsev VV, Noodleman L, Sharpless KB, et al.
J Am Chem Soc 2005;127:210e6.
References
[30] Gujadhur R, Venkataraman D, Kintigh JT. Tetrahedron Lett 2001;42:4791e3.
[31] Martinez-Castro N, Magenau AJD, Storey RF. J Polym Sci Polym Chem 2010;
48(12):2533e45.
[32] Lipshutz BH, Taft BR. Angew Chem Int Ed 2006;45(48):8235e8.
[33] Nulwala H, Takizawa K, Odukale A, Khan A, Thibault RJ, Taft BR, et al.
Macromolecules 2009;42(16):6068e74.
[34] Sheng X, Mauldin TC, Kessler MR. Polymer 2010;48(18):4093e102.
[35] As a control test to ensure that there are not metal contaminants inadvertedly
catalyzing the “uncatalyzed” reactions (sample 1), we thrice ran each of the
alkyne and azide monomers through a column of basic alumina to remove any
potential trace metal contaminants. Subsequent DSC curing tests with the
purified monomers yielded exothermic peaks nearly identical in position and
magnitude to those presented in the original manuscript (i.e. with monomers
not run through columns), indicating that no metal contaminants exist in the
monomers.
[1] (a) Huisgen R. Angew Chem 1963;75:742e54;
(b) Huisgen R. Angew Chem 1963;75:604e37;
(c) Huisgen R. Chemisch Weekblad 1963;59:89e101;
(d) Huisgen R, Szeimies G, Moebius L. Chemische Berichte 1967;100:
2494e507;
(e) Huisgen R, Von Fraunberg K, Sturm HJ. Tetrahedron Lett 1969;30:
2589e94.
[2] (a) Kolb HC, Finn MG, Sharpless KB. Angew Chem Int Ed 2001;40:2004e21;
(b) Gil MV, Arévalo MJ, López Ó. Synthesis 2007;11:1589e620.
[3] Rostovtsev VV, Green LG, Fokin VV, Sharpless KB. Angew Chem Int Ed 2002;
41(14):2596e9.
[4] Tornoe CW, Christensen C, Meldal M. J Org Chem 2002;67(9):3057e64.
[5] Opsteen JA, Hest van JCM. Chem Commun; 2005:57e9.
[6] Klaikherd A, Ghosh S, Thayumanavan S. Macromolecules 2007;40(24):
8518e20.
[36] Gragert M, Schunack M, Binder WH. Macromol Rapid Commun 2011;32:
419e25.
[7] Durmaz H, Dag A, Altintas O, Erdogan T, Hizal G, Tunca U. Macromolecules
2007;40(2):191e8.
[37] Kissinger HE. Anal Chem 1957;29(11):1702e6.
[8] Lin W, Fu Q, Zhang Y, Huang J. Macromolecules 2008;41(12):4127e35.
[9] Laurent BA, Grayson SM. J Am Chem Soc 2006;128(13):4238e9.
[10] Tsarevsky NV, Sumerlin BS, Matyjaszewski K. Macromolecules 2005;38(9):
3558e61.
[38] (a) Yousefi A, Lafleur PG, Gauvin R. Polym Compos 1997;18:157e69;
(b) Ganguli S, Dean D, Jordan K, Price G, Vaia R. Polymer 2003;44(22):
6901e11;
(c) Simon SL, Gillham JK. J Appl Polym Sci 1993;47(3):461e85.
[39] Menard KP. Dynamic mechanic analysis. CRC Press LLC; 1999.
[11] Vogt AP, Sumerlin BS. Macromolecules 2006;39(16):5286e92.
[12] Gao H, Matyjaszewski K. J Am Chem Soc 2007;129(20):6633e9.