Effect of crosslinker multiplicity on the gel point in ATRP

Article abstract of DOI:10.1002/pola.23970

The experimental gelation was studied in atom transfer radical polymerization (ATRP) of methyl acrylate (MA) with various branching reagents that structurally differ by the number of vinyl groups (multiplicity). MA was copolymerized with branching reagent

Full text of DOI:10.1002/pola.23970

Effect of Crosslinker Multiplicity on the Gel Point in ATRP
WIM VAN CAMP,1,2 HAIFENG GAO,1 FILIP E. DU PREZ,2 KRZYSZTOF MATYJASZEWSKI1
1Department of Chemistry, Center for Macromolecular Engineering, Carnegie Mellon University,
4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213
2Polymer Chemistry Research Group, Department of Organic Chemistry, Ghent University,
Krijgslaan 281 S4, Ghent B-9000, Belgium
Received 2 February 2010; accepted 4 February 2010
DOI: 10.1002/pola.23970
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: The experimental gelation was studied in atom
transfer radical polymerization (ATRP) of methyl acrylate (MA)
with various branching reagents that structurally differ by the
number of vinyl groups (multiplicity). MA was copolymerized
with branching reagents containing 2, 3, 4, 5, or 6 acrylate moi-
eties per molecule, respectively. Reactions with a constant con-
centration of branching vinyl groups (the same molar ratio of
[Vinylbranch]0/[Initiator]0) revealed a different gelation behavior
when ethylene glycol diacrylate (2A) and trimethylolpropane tri-
acrylate (3A) were used as crosslinkers, whereas the reactions
using pentaerythritol tetraacrylate (4A) showed similar gelation
behavior as compared to reactions using 3A. Additional reac-
tions with dipentaerythritol pentaacrylate (5A) and hexafunc-
tional acrylate crosslinkers (6A) revealed the presence of a
concentration-dependant gelation behavior. On the other hand,
in reactions with the same molar concentration of various cross-
linkers, gelation occurred at progressively lower MA conver-
sions for reactions with an increasing number of vinyl groups
per crosslinker. In addition, the number of unreacted pendant
vinyl groups in the sols was compared for reactions with
different multiplicity of the crosslinker and various ratios of
[Vinylbranch]0/[Initiator]0. Finally, a linear oligomeric crosslinker
containing multiple branching vinyl groups along the side chain
was used as the branching reagent for gelation. The gelation
behavior during the ATRP of MA with the linear crosslinker was
different as compared to the use of the aforementioned cross-
C
V
linkers with a star-like architecture.
2010 Wiley Periodicals,
Inc. J Polym Sci Part A: Polym Chem 48: 2016–2023, 2010
KEYWORDS: ATRP; crosslinker; gelation; multiplicity
INTRODUCTION Conventional radical polymerization (RP) of
a monovinyl monomer with a small amount of divinyl cross-
linker is a widespread method for the synthesis of branched
polymers or gels. However, the slow and continuous initia-
tion process, fast chain propagation, and termination reac-
tions that are inherent to conventional RP result in networks
with an inhomogeneous and uncontrolled structure.1–4
Recently, several research groups have shown that the appli-
cation of controlled radical polymerization (CRP) techni-
ques5–9 produces gels with a more homogeneous structure,
as a result of the fast initiation process and reversible deacti-
vation of propagating polymer chains.10–13 The fast initiation
reactions, relative to propagation, result in a quick conver-
sion of nearly all initiators to growing primary chains. On
the other hand, the dynamic equilibrium between the active
species, bearing a radical at the polymer chain end, and the
dormant species ensures a low steady concentration of radi-
cals. During each activation/deactivation cycle, only a few
monomers are added to the growing polymer chain before it
is quickly deactivated to the dormant state. The dormant
state allows for chain relaxation and diffusion of reagents. As
a result, the probability of reaction of each vinyl species,
monomer, ‘‘unreacted’’ crosslinker or pendant vinyl group, is
dependent on their relative concentration and reactivity.
Therefore, the branched polymers or gels have a more homo-
geneous distribution of branching points than the polymers
synthesized by RP methods at similar reaction conditions.11
Recently, several research groups systematically studied the
synthesis of branched polymers and/or gels by the copoly-
merization of monovinyl monomer and divinyl crosslinker
using atom transfer radical polymerization (ATRP).10,14–25
Sherrington20,21 and Armes18,19,22 focused on the synthesis
of branched polymers based on copolymerization of metha-
crylate and dimethacrylate crosslinker. Zhu and coworkers
studied the homopolymerization25 of ethylene glycol dime-
thacrylate and copolymerization23,24 of methyl methacrylate
and dimethacrylate crosslinker using bulk ATRP. Our group
applied the ATRP of methyl acrylate (MA) and ethylene gly-
col diacrylate as a model system to study the dependence of
the experimental gel points on various parameters, because
of its fast initiation process, high initiation efficiency, and
low polydispersity in all performed experiments.14 The effect
of several parameters was studied, for example, the molar
ratio of divinyl crosslinker to initiator,14 the concentration of
Correspondence to: K. Matyjaszewski (E-mail: km3b@andrew.cmu.edu)
C
V
Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 48, 2016–2023 (2010) 2010 Wiley Periodicals, Inc.
2016
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
SCHEME 1 Synthesis of polymer
networks by ATRP of MA and
different branching reagents.
all reagents during the gelation process,15 the relative reac-
tivity of divinyl crosslinker compared to monomer,16 and the
initiation efficiency and polydispersity of the primary
chains.17 Furthermore, the experimental results were com-
pared with values obtained from Monte Carlo simulations.26
Finally, an AB* inimer (initiator-monomer containing vinyl
bond A and initiator fragment B* in one molecule) was also
grade, Aldrich) was purified via silica liquid chromatography
with mixture of hexane/ethyl acetate 7/3 by volume as the
eluent. Dipentaerythritol penta-/hexa-acrylate (5A, technical
grade, Aldrich) was determined to have an average multiplic-
ity of 5 by 1H NMR, and was used as received. CuBr (98%,
Acros) was purified using a modified literature procedure.28
explored as
a branching reagent for the synthesis of
branched copolymers and gels by copolymerization with
monomer and divinyl crosslinker using ATRP,27 and the
results were compared with gelation reactions using divinyl
and trivinyl crosslinker.
In this article, we aim to expand the systematic study on the
gelation in ATRP and investigate the effect of crosslinker
multiplicity, that is, the number of vinyl groups per cross-
linker molecule, on the experimental gel points. Copolymer-
ization of MA was carried out with various branching
reagents: ethylene glycol diacrylate (2A), trimethylolpropane
triacrylate (3A), pentaerythritol tetraacrylate (4A), dipentaery-
thritol pentaacrylate (5A), and a hexafunctional acrylate (6A)
crosslinker (see Schemes 1 and 2). To keep the system com-
parable to those previously studied, the experimental gel
points based on the monomer (MA) conversions were deter-
mined and systematically compared to each other for reac-
tions where the molar ratios of branching vinyl groups to ini-
tiator ([Vinylbranch]0/[Initiator]0) and the crosslinker species
were varied. The effect of crosslinker multiplicity was also
investigated in a series of reactions where the molar concen-
tration of the various crosslinkers was kept constant. In addi-
tion, the behavior of the aforementioned star-like crosslinkers
(3A, 4A, 5A, 6A) was compared with a linear oligomeric multi-
vinyl crosslinker molecule, containing an average of seven
dangling vinyl groups along the polymer chain.
EXPERIMENTAL
Materials
Methyl acrylate (MA, 99%), 2-hydroxyethyl acrylate (HEA,
96%), n-butyl acrylate (nBA, 99%), and ethylene glycol dia-
crylate (2A, 90%) were purchased from Aldrich and purified
twice by passing through a column filled with basic alumina
to remove the inhibitor. The trivinyl crosslinker, trimethylol-
propane triacrylate (3A, technical grade, Aldrich) and tetra-
vinyl crosslinker, pentaerythritol tetraacrylate (4A, technical
SCHEME 2 Synthesis of hexafunctional acrylate crosslinker
(6A).
EFFECT OF CROSSLINKER MULTIPLICITY, VAN CAMP ET AL.
2017

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
All other reagents: methyl 2-bromopropionate (MBP), ethyl
2-bromopropionate (EBrP), N,N,N0,N,00N00-pentamethyldiethy-
lenetriamine (PMDETA), acryloyl chloride, N,N0-dicyclohexyl-
carbodiimide (DCC), 4-(dimethylamino)pyridine, p-toluene-
sulfonic acid monohydrate, 2,2-bis(hydroxymethyl)propionic
acid, 2,2-dimethoxypropane, triethylamine, 1,1,1-tris(4-
hydroxyphenyl)ethane, 4-Hydroxy-TEMPO, CuBr2, and sol-
vents were purchased from Aldrich with the highest purity
and used as received without further purification.
purification, the product was dissolved in a mixture of 30
mL of THF and 30 mL of 1 M HCl (aq.). The reaction mixture
was stirred for 2 h before the precipitated product was fil-
tered and washed with water. The product was dried in vac-
ꢁ
uum at 50 C, yielding a white solid. Yield: 3.83 g (90%).
1H NMR (acetone-d6): d ¼ 1.31 (s, 9H, ACH3), 2.21 (s, 3H,
ACH3), 3.76–3.89 (m, 12H, AC(CH2OA)2), 7.06–7.17 (m,
12H, AOPhA).
Esterification with Acryloyl Chloride
Characterization
2 (1.52 g, 2.32 mmol, 13.9 mmol OH groups), 25 mL dry
dichloromethane, triethylamine (2.81 g, 3.88 mL, 27.9
mmol), and 10.0 mg of radical inhibitor 4-hydroxy-TEMPO
were added to a clean, dry, round-bottom flask. The flask
was placed in a an ice-water bath and the mixture was
Monomer conversions were determined from the concentra-
tion of the unreacted monomer in the samples periodically
removed from the reactions using a Shimadzu GC-17A gas
chromatograph, equipped with a capillary column (DB-Wax,
30 m ꢀ 0.54 mm ꢀ 0.5 lm, J and W Scientific). DMF was
used as internal standard for calculation of monomer conver-
sions. After filtration through 220 nm PTFE filter, the poly-
mer samples were separated by GPC [Polymer Standards
Services (PSS) columns (guard, 105, 103, and 102 Å), with
ꢁ
cooled down to 0 C. Acryloyl chloride (2.52 g, 2.27 mL, 27.8
mmol, 2 eq. to OH groups) was added dropwise to the reac-
tion mixture over a period of 20 min. After 1 h, the flask
was removed from the ice-water bath and the reaction mix-
ture was stirred for 24 h at room temperature. During this
period, the reaction mixture changed from a yellowish color
to brown. The reaction mixture was washed successively
with 100 mL of 1 M HCl, 100 mL of 1 M NaOH, and 100 mL
of deionized water and then dried over anhydrous MgSO4 for
1 h. Ten milligram of radical inhibitor 4-hydroxy-TEMPO was
added to the mixture and the solvent was removed via
rotary evaporation and oil pump vacuum. The final product
was a viscous oil.
ꢁ
THF eluent at 35 C, flow rate ¼ 1.00 mL/min and differen-
tial refractive index (RI) detector (Waters, 2410)]. The appa-
rent molecular weights and polydispersities (Mw/Mn) were
determined with a calibration based on linear poly(methyl
methacrylate) (polyMMA) standards using WinGPC 6.0 soft-
ware from PSS. The GPC curves obtained with flat baseline
were imported into the WinGPC software for calculation of
their apparent molecular weights and polydispersity. 1H
NMR spectra, using CDCl3 or acetone-d6 as solvent,
were measured on a Bruker Avance 300 MHz spectrometer
1
The structure was verified by H NMR spectroscopy (CDCl3):
ꢁ
d ¼ 1.44 (s, 9H, ACH3), 2.13 (s, 3H, ACH3), 4.47 (s, 12H,
AC(CH2OA)2), 5.85–5.89 (2d, 1H, ACH2CHC(O)O), 6.08–6.18
(dd, 1H, ACH2CHC(O)OA), 6.40–6.46 (2d, 1H, ACH2CHC(O)O),
6.91–7.08 (m, 12H, AOPhA).
at 27 C.
Synthesis of Hexafunctional Acrylate (6A) Crosslinker
Synthesis of 2,2-Bis((2,2-propyl)dioxymethyl)
propionic acid 1
Synthesis of PolyMA-Based Gels by ATRP of MA
with 2A, 3A, 4A, 5A, or 6A
2,2-Bis(hydroxymethyl)propionic acid (bisMPA) (10.0 g, 74.6
mmol), was reacted with 2,2-dimethoxypropane (13.8 mL,
112 mmol, 1.5 eq. to bisMPA), and p-TSA (0.71 g, 5 mol % to
bisMPA) in 50 mL of acetone at room temperature. After 2 h,
1 mL of NH4OH (aqueous, 30%)/EtOH (1/1) solution was
added to neutralize the catalyst. The reaction mixture was
concentrated by rotary evaporation, the residue was redis-
solved in CH2Cl2, and extracted twice with 20 mL of H2O.
The organic phase was dried over MgSO4, filtered, and
evaporated to yield 8.3 g (63%) of white crystals.
A typical procedure for the ATRP of MA and 2A is briefly
described, starting with a ratio of reagents [MA]0/[2A]0/
[EBrP]0/[CuBr]0/[CuBr2]0/[PMDETA]0 ¼ 50/2/1/0.45/0.05/
0.5, [MA]0 ¼ 6.0 M. A clean and dry Schlenk flask was
charged with MA (3.50 mL, 0.039 mol), 2A (0.290 mL, 1.56
mmol), PMDETA (81.2 lL, 0.39 mmol), and N,N-dimethylfor-
mamide (DMF, 2.51 mL). The flask was deoxygenated by five
freeze-pump-thaw cycles. During the final cycle, the flask
was filled with nitrogen before CuBr (50.2 mg, 0.35 mmol)
and CuBr2 (8.7 mg, 0.039 mmol) were quickly added to the
frozen mixture. No precautions were taken to avoid moisture
condensation. The flask was sealed with a glass stopper then
evacuated and back-filled with nitrogen five times before it
1H NMR (CDCl3): d ¼ 1.20 (s, 3H, ACH3), 1.41–1.44 (d, 6H,
AC(CH3)2), 3.65–4.20 (dd, 4H, AC(CH2OA)2).
Synthesis and Deprotection of 2
1,1,1-Tris(4-hydroxyphenyl)ethane (3.73 g, 12.2 mmol) and 1
(7.0 g, 40.2 mmol, 1.1 eq. to OH groups) was dissolved in 25
mL of dry CH2Cl2. N,N0-dicyclohexylcarbodiimide (DCC) (11.3
g, 1.5 eq. to OH groups) and 4-(dimethylamino)pyridine
(1.13 g, 10 wt % to DCC) was added. The mixture was
stirred overnight at room temperature. The formed urea was
removed by filtration. The filtrate was concentrated and the
resulting viscous oil was purified by silica gel column chro-
matography (eluens hexane/ethyl acetate 3/2). The pure
fractions were concentrated, resulting in a viscous oil. After
ꢁ
was immersed in an oil bath at 60 C. Four cylindrical glass
tubes were initially added into the reaction system. EBrP ini-
tiator (0.101 mL, 0.78 mmol) initiator was added to start the
gelation reaction. At timed intervals, samples were with-
drawn via a syringe for GC measurements of monomer con-
version. The system formed a gel at certain moment when
the reaction fluid lost its mobility at an upside down posi-
tion for 10 s. After gelation, the reaction was kept at 60 ꢁC
for another 2 days before stopping the reaction via exposure
2018
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
TABLE 1 Comparison of ATRP Reactions of MA Using
to air. The catalyst was removed from the gel by repeated
THF extraction and filtration. The experimental procedures
were similar for ATRP of MA with 3A, 4A, 5A, or 6A.
Crosslinkers with Different Multiplicitya
c
ConvMA,gel
Synthesis of Copolymer Poly(HEA-co-nBA)
b
[Vinylbranch]0/[Initiator]0
2A
3A
4A
5A
6A
A polymerization was conducted using reaction conditions
[nBA]0/[HEA]0/[MBP]/[CuBr]0/[CuBr2]0/[PMDETA]0 ¼ 25/
25/1/0.45/0.05/0.5 (in 25 vol % of DMF). nBA (41.7 mmol,
6.0 mL), HEA (41.7 mol, 4.89 mL), Cu(II)Br2 (0.083 mmol,
18.6 mg), PMDETA as ligand (0.75 mmol, 0.174 mL), and
DMF as solvent (3.63 mL, 25 vol %) were added to a reac-
tion flask and the mixture was bubbled with N2 for 1 h to
remove oxygen from the reaction mixture. After that, Cu(I)Br
(0.749 mmol, 0.108 g) was added and the reaction flask was
placed in an oil bath at 50 ꢁC. When the reaction mixture
reached the desired reaction temperature, the polymerization
was started by adding MBP (1.66 mmol, 0.186 mL) as the
initiator. Samples were withdrawn periodically to monitor
the monomer conversion (by GC) and the average molecular
weight (by GPC). The reaction was ended after 1800 by cool-
ing the reaction mixture in liquid nitrogen. GC analysis
revealed a monomer conversion of 41% for HEA and 34%
for nBA, respectively. This ratio was confirmed by 1H NMR
analysis of the pure polymer. The copolymer was separated
from the monomer by precipitation in a large excess of cold
distilled water, followed by centrifugation. After drying in
vacuum at room temperature overnight, the pure copolymer
was obtained as a highly viscous material.
d
2.20
3.00
4.00
10.0
a
0.97
0.86
0.79
0.48
0.94
0.78
0.67
0.40
0.89
0.79
0.67
0.39
0.99
0.83
0.65
0.34
/
d
/
0.82
0.34
Experimental conditions: (1) ‘‘2A’’ reaction: ATRP of MA and 2A with
[MA]0/[2A]0/[EBrP]0/[CuBr]0/[CuBr2]0/[PMDETA]0 50/A/1/0.45/0.05/0.5;
(2) ‘‘3A’’ reaction: ATRP of MA and 3A: [MA]0/[3A]0/[EBrP]0/[CuBr]0/
[CuBr2]0/[PMDETA]0 50/B/1/0.45/0.05/0.5; (3) ATRP of MA and 4A:
[MA]0/[3A]0/[EBrP]0/[CuBr]0/[CuBr2]0/[PMDETA]0 50/C/1/0.45/0.05/0.5;
(4) ATRP of MA and 5A: [MA]0/[5A]0/[EBrP]0/[CuBr]0/[CuBr2]0/[PMDETA]0
50/D/1/0.45/0.05/0.5; (5) ATRP of MA and 6A: [MA]0/[6A]0/[EBrP]0/
50/E/1/0.45/0.05/0.5. All reactions were
¼
¼
¼
¼
[CuBr]0/[CuBr2]0/[PMDETA]0
¼
performed with [MA]0 ¼ 6.0 M in DMF at 60 ꢁC and stopped after 2
days to reach the complete conversions.
b
[Vinylbranch]0/[Initiator]0 ¼ (2 ꢀ A) for ‘‘2A’’ reaction, (3 ꢀ B) for ‘‘3A’’
reaction, (4 ꢀ C) for reaction with 4A, (5 ꢀ D) for reaction with 5A, and
(6 ꢀ E) for reaction with 6A. For the comparable reactions with the
same value of [Vinylbranch]0/[Initiator]0, B ¼ 2 ꢀ A/3, C ¼ A/2, D ¼ 2 ꢀ A/
5, E ¼ A/3.
c
Conversions of MA immediately before gelation. The experimental gel
point was the moment when the reaction fluid lost its mobility when
held at an upside down position for 10 s.
d
No gelation occurred under these conditions.
tion, the effect of crosslinker multiplicity was investigated in
a series of reactions where the molar concentration of the
various crosslinkers was kept at a constant value.
Synthesis of Multivinyl Crosslinker Based
on Poly(HEA-co-nBA)
Poly(HEA-co-nBA) copolymer (Mn ¼ 1700 g/mol and Mw/Mn
¼ 1.11) was reacted with acryloyl chloride in a similar pro-
cedure as described above, 0.1 mol % of radical inhibitor
4-hydroxy-TEMPO was added relative to vinyl groups to
prevent crosslinking during the esterification reaction. The
reaction product was purified by dialysis in THF, yielding a
polymer with pendant vinyl groups (Mn ¼ 1850 g/mol and
Mw/Mn ¼ 1.12). 1H NMR spectroscopy of the pure polymer
revealed an esterification yield of about 90%.
Gelation Reactions with Constant Molar Ratio
of [Vinylbranch]0/[Initiator]0
In the first series of reactions, gelation reactions with ethyl-
ene glycol diacrylate (2A) and trimethylolpropane triacrylate
(3A) crosslinkers were compared. The molar ratio of vinyl
groups that can introduce branching (2 per molecule for 2A;
3 for 3A, etc.) to initiator was kept at a constant value (con-
stant [Vinylbranch]0/[Initiator]0). For various molar ratios of
[Vinylbranch]0/[Initiator]0 ranging from 2.20 to 10, experi-
mental gelation occurred consistently at lower MA conver-
sions for reactions with the trivinyl crosslinker 3A, as com-
pared to the use of divinyl crosslinker (Table 1). Upon
incorporation of a trivinyl crosslinker into a growing poly-
mer chain, two pendant vinyl groups are generated. If one of
them is consumed by intramolecular cyclization, the second
one is still available to react and form an intermolecular
crosslinkage. In contrast, if the pendant vinyl group from an
incorporated divinyl crosslinker reacts intramolecularly, this
crosslinker does not contribute to the increase of the molec-
ular weight of the branched polymers. As a result, the use of
a trivinyl crosslinker for the copolymerization leads to a net-
work with a higher extent of intermolecular crosslinkages
than the use of 1.5 eq. of divinyl crosslinker.
RESULTS AND DISCUSSION
The ATRP of MA monomer and various crosslinkers contain-
ing different numbers of acrylate moieties was performed
using CuBr/PMDETA as the catalyst (10 mol % of CuBr2 of
total copper species was preadded in the system), ethyl-2-
bromopropionate (EBrP) as the initiator with an initial molar
ratio of [MA]0/[EBrP]0 ¼ 50 in DMF ([MA]0 ¼ 6.0 M) at 60
ꢁC. These reaction conditions have been previously reported
as a good system for ATRP of acrylate-based monomers,
exhibiting fast initiation kinetics, a high initiation efficiency,
and controlled polymer chain growth with low polydispersity
of the primary chains.14
In this article, two approaches were used to investigate the
effect of crosslinker multiplicity on experimental gelation.
First, the concentration of branching vinyl groups (molar ra-
tio of [Vinylbranch]0/[Initiator]0) was kept constant for series
of reactions with increasing crosslinker multiplicity. In addi-
In similar series of reactions with various molar ratios of
[Vinylbranch]0/[Initiator]0, gelation reactions were also con-
ducted using a tetravinyl crosslinker (pentaerythritol tetraa-
crylate, 4A),
a
pentavinyl crosslinker (dipentaerythritol
EFFECT OF CROSSLINKER MULTIPLICITY, VAN CAMP ET AL.
2019

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
multiplicity of the crosslinker. This is schematically depicted
in Figure 2.
For low [Vinylbranch]0/[Initiator]0, the crosslinker concentra-
tion may drop below a critical value, resulting in a more pro-
nounced intramolecular cyclization as the effect of a lower
concentration in the reaction medium. It was previously
reported that a decreasing concentration of the reaction me-
dium results in delayed experimental gelation due to signifi-
cantly increased intramolecular cyclization.15 Similarly, a
decreased crosslinker concentration, originating from the
increased multiplicity of the crosslinker with a constant ratio
of [Vinylbranch]0/[Initiator]0, could also lead to delayed exper-
imental gelation. In other words, an increasing multipli-
city increased the local concentration of pendant vinyl
groups within one chain and increased the intramolecular
cyclization.
FIGURE 1 Conversion of MA at gelation for ATRP of MA and
crosslinker as a function of crosslinker multiplicity for various
constant molar ratios of [Vinylbranch]0/[Initiator]0.
On the other hand, if the [Vinylbranch]0/[Initiator]0 ratio is
rather high, the crosslinker concentration may be sufficient
to prevent an intramolecular cyclization, resulting in earlier
gelation with increasing crosslinker multiplicity.
pentaacrylate, 5A), and a hexafunctional acrylate crosslinker
(6A). It should be noted that 5A crosslinker is based on
dipentaerythritol penta-/hexa-acrylate, which could contain
small amounts of triester and hexafunctional compound. In
our study, no further purification was performed as the aver-
Gelation Reactions with Constant Molar
Concentration of Crosslinker
The previously discussed reactions with
a
constant
[Vinylbranch]0/[Initiator]0 ratio correlated the experimental
gelation behavior with a dilution effect, resulting from an
increasing multiplicity of the crosslinker. However, gelation
reactions with a relatively high [Vinylbranch]0/[Initiator]0 ratio
still showed gelation at lower monomer conversion for
increasing multiplicity of the crosslinker. To further investi-
gate the influence of the crosslinker multiplicity in relation
with its concentration, two series of gelation reactions were
performed with two constant molar ratios of [monomer]/
[initiator]/[crosslinker] ([M]0/[I]0/[X]0) ¼ 50/1/1 and 50/1/
2, under similar reaction conditions as described above. In
other words, the molar concentration of the crosslinker in
the medium is identical for all gelation reactions, regardless
of the number of branching vinyl groups per crosslinker mol-
ecule (see schematic depiction in Fig. 3).
1
age multiplicity was determined by H NMR as being a pen-
tafunctional acrylate. The synthesis of hexafunctional acry-
late 6A is depicted in Scheme 2. The results of the gelation
reactions are summarized in Table 1.
For reactions using the trivinyl crosslinker 3A, gelation
occurred at consistently lower monomer conversion as com-
pared to the gelation by using 2A crosslinker (Fig. 1). How-
ever, for reactions using the tetravinyl crosslinker 4A, no sig-
nificant influence of the crosslinker multiplicity was
observed under the same [Vinylbranch]0/[Initiator]0 ratio, as
compared to the gelation by using 3A crosslinker. When the
crosslinker containing five acrylate moieties per molecule
was used, a different trend was observed in comparable
reactions. For low [Vinylbranch]0/[Initiator]0 ratios (i.e., 2.2
and 3), gelation seems to be delayed with increasing multi-
plicity of the crosslinker. For instance, for the reaction with
5A, the experimental gel point was only observed at 99%
conversion for a ratio of [Vinylbranch]0/[Initiator]0 ratio ¼
2.2. In contrast, for higher [Vinylbranch]0/[Initiator]0 ratios
(i.e., 4 and 10), increased multiplicity of the crosslinker still
resulted in gelation at earlier MA conversion. The effect
tends to be even more clear for reactions with [Vinylbranch]0/
[Initiator]0 ¼ 10. This trend is confirmed for the reactions
using hexafunctional crosslinker 6A. In this case, no gelation
was observed for reactions with a [Vinylbranch]0/[Initiator]0
ratio ¼ 2.2 and 3, whereas gelation was significantly delayed
for the reaction with a [Vinylbranch]0/[Initiator]0 ratio ¼ 4.
Only for the reaction with a [Vinylbranch]0/[Initiator]0 ratio ¼
10, no gelation delay was observed. We attribute these
observations to a dilution effect: a constant [Vinylbranch]0/
[Initiator]0 ratio directly implies that the concentration of
crosslinker in the reaction medium decreases with increasing
In contrast to reactions with a constant molar ratio of [Vinyl-
branch]0/[Initiator]0 for the various crosslinkers, this study
avoids the influence of effects that are directly related to
dilution of the crosslinker molecules in the reaction medium.
For example, for the series of reactions with [M]0/[I]0/[X]0
¼ 50/1/1, the [Vinylbranch]0/[Initiator]0 ratio was 2, 3, 4, 5,
and 6, for divinyl crosslinker 2A, trivinyl crosslinker 3A, tet-
ravinyl crosslinker 4A, pentavinyl crosslinker 5A, and hexa-
vinyl crosslinker 6A, respectively. At [Vinylbranch]0/[Initiator]0
¼ 2, no gelation was observed when 2A was used as cross-
linker. The experimental gel points for these series of gela-
tion reactions with constant concentration of crosslinker are
displayed in Figure 4. For a constant concentration of the
crosslinker, gelation occurred consistently at lower MA con-
versions for reactions with an increasing number of vinyl
groups per crosslinker molecule.
A similar trend was
observed for a series of gelation reactions with [M]0/[I]0/
[X]0 ¼ 50/1/2.
2020
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
FIGURE 2 Schematic depiction of reaction medium of ATRP of MA (blue dots) and various crosslinkers with different multiplicity
(in red: 2A, 3A, 4A, 5A, and 6A containing 2, 3, 4, 5, or 6 vinyl groups per molecule, respectively) for a series of gelation reactions
with constant molar ratio of [Vinylbranch]0/[Initiator]0 ¼ 12.
Determination of Pendant Vinyl Groups in the Sols
The reactions with various constant ratios of [Vinylbranch]0/
[Initiator]0 for various crosslinkers with different multiplicity
suggested a gelation behavior related to the concentration of
the crosslinker in the reaction medium. To further investi-
gate this effect, the number of unreacted pendant vinyl
groups in the sols was determined for reactions with differ-
ent multiplicity of the crosslinker and various ratios of
[Vinylbranch]0/[Initiator]0. If equal reactivity of each vinyl
group is assumed, the fraction of each species originating
from the multivinyl crosslinker is determined by the statisti-
cal combination formula below.
of the pendant vinyl groups in the sols (fA,p) during ATRP of
monomer and trivinyl crosslinker is the sum of the pendant
vinyl groups present in the fractions of k ¼ 1 and k ¼ 2.
The theoretical and experimental fractions of pendant vinyl
groups in the sols (fA,p) as a function of MA conversion dur-
ing ATRP of MA and pentavinyl crosslinker 5A were com-
pared for a few reactions with various molar ratios of [Vinyl-
branch]0/[Initiator]0 (Fig. 5). The dots in Figure 5 represent
1
the experimental results of fA,p determined by H NMR anal-
ysis of the sols just before gelation. Unreacted MA and free
crosslinker were removed from the polymer samples by
repeated precipitation/redissolving in hexane/tetrahydrofu-
ran. 1H NMR integration results determined the molar ratio
of pendant vinyl groups to reacted MA units in the polymer
backbone, which was further converted into the fA,p taking
into account the MA conversion.14 The results in Figure 5
show that the experimental fA,p was in good agreement with
the calculated fA,p for reactions with relatively low ratio of
[Vinylbranch]0/[Initiator]0 (i.c. 3 and 4). However, the experi-
mental value of fA,p was higher than the theoretical value
when the ratio of [Vinylbranch]0/[Initiator]0 was 10. This dif-
ference could be attributed to the steric protection of dan-
gling chains, resulting in the pendant vinyl groups less likely
to react. For a relatively low [Vinylbranch]0/[Initiator]0 ratio
(3 and 4), the distance between the various crosslinkages
would be relatively larger (lower crosslink density), com-
pared to reactions with a higher [Vinylbranch]0/[Initiator]0 ra-
tio (6 and 10; higher crosslink density). Intermolecular
crosslinking reactions would be more difficult in the latter
case, resulting in a higher fA,p than the theoretical value.
n
nꢂk
PðK ¼ kÞ ¼ ðkÞpkð1 ꢂ pÞ
n!
n
with ðkÞ ¼
k!ðn ꢂ kÞ!
where p is conversion of M and k ¼ number of vinyl groups
reacted out of total number n.
This formula expresses the probability that k (out of total
number n) vinyl groups have reacted at monomer conversion
p, and defines the fraction of each species relative to the
starting amount of unreacted crosslinker. For a crosslinker
containing n vinyl groups, n þ 1 species can be distin-
guished. For example, four different species originating from
the crosslinker can be observed in the ATRP of monomer
and 3A crosslinker: (i) a fraction with none of the vinyl
groups reacted (k ¼ 0), representing the unreacted free
crosslinker; (ii) a fraction of the crosslinker with one vinyl
group reacted (k ¼ 1), which is incorporated in a primary
chain and contains two pendant vinyl groups; (iii) a fraction
of the crosslinker with two vinyl groups reacted (k ¼ 2) and
1 pendant vinyl group; and (iv) a fraction of crosslinker that
has fully reacted (k ¼ 3). Therefore, the theoretical fraction
Further investigation for a molar ratio of [Vinylbranch]0/[Ini-
tiator]0 ¼ 10 was done for trivinyl crosslinker 3A, and tetra-
vinyl crosslinker 4A. It was previously reported that theoreti-
cal and experimental fA,p values for ATRP of MA and 2A
FIGURE 3 Schematic depiction of reaction medium of ATRP of MA (blue dots) and various crosslinkers with different multiplicity
(in red: 2A, 3A, 4A, 5A, and 6A containing 2, 3, 4, 5, or 6 vinyl groups per crosslinker molecule, respectively) for a series of gelation
reactions with a constant concentration of crosslinker.
EFFECT OF CROSSLINKER MULTIPLICITY, VAN CAMP ET AL.
2021

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
FIGURE 6 Theoretical and experimental fraction of pendant ac-
rylate groups (fA,p) in the sols as a function of crosslinker mul-
tiplicity for ATRP of MA and 3A, 4A, and 5A, respectively, for
[Vinylbranch]0/[Initiator]0 ratio ¼ 10. Samples were taken close to
FIGURE 4 Conversion of MA at experimental gel point for
ATRP of MA and crosslinker with various multiplicities, using a
constant molar crosslinker concentration [X].
were in a good agreement.14,26 For relatively low [Vinyl-
branch]0/[Initiator]0 ratio (i.c. 3 and 4), no significant differen-
ces between the theoretical and experimental fA,p for ATRP
of MA and 3A and 4A were observed, just as in the case of
ATRP of MA and 5A. However, for a high ratio of [Vinyl-
branch]0/[Initiator]0 (i.c. 10), a growing discrepancy between
the theoretical and experimental fA,p was found with increas-
ing multiplicity of the crosslinker (see Fig. 6).
gelation up to 91% MA conversion, in contrast to reactions
with 5A (experimental gelation at 65% MA conversion) and
6A (experimental gelation at 82% MA conversion). In the
reactions with a constant concentration of crosslinker ([M]0/
[I]0/[X]0 ¼ 50/1/1), ATRP of MA and 7A showed a delayed
gelation (at 99% MA conversion) in comparison to a similar
reaction with the same concentration of 5A (gelation at 54%
MA conversion). The difference in the gelation behavior
could be attributed to the oligomeric nature of the multivinyl
crosslinker and also to the linear structure of 7A. The
acrylate crosslinking groups located next to each other along
the backbone could facilitate intramolecular crosslinking
reactions.
Gelation with a Polymeric Multivinyl Crosslinker (7A)
In addition to the use of the crosslinkers with well-defined
number of vinyl groups per molecule, we have also studied a
gelation in the presence of an oligomeric (polymer-based)
crosslinker. The crosslinker was prepared based on a random
copolymer of n-butyl acrylate (nBA) and 2-hydroxyethyl ac-
rylate (HEA), followed by the esterification of the hydroxyl
groups in the side chain to acrylate moieties using acryloyl
chloride. The average number of acrylate moieties per macro-
molecule was 7, as determined by 1H NMR. In the reactions
with a constant molar ratio of [Vinylbranch]0/[Initiator]0 ¼ 4,
the ATRP of MA and multivinyl crosslinker 7A showed no
CONCLUSIONS
The effect of various types of crosslinkers containing differ-
ent numbers of vinyl groups per molecule (2, 3, 4, 5, or 6)
on the gelation behavior was studied during the ATRP of
MA. At a constant ratio of branching vinyl groups to initiator,
a concentration-dependant gelation for crosslinkers with dif-
ferent numbers of vinyl groups per molecule was observed.
At a low targeted crosslink density, the experimental gelation
was delayed for crosslinkers with higher number of vinyl
groups per molecule (5 or 6). At a higher targeted crosslink
density, increasing the number of vinyl groups per cross-
linker molecule led to slightly accelerated gelation. When a
constant molar concentration of various crosslinkers was
used, experimental gelation occurred consistently at lower
monomer conversions for reactions with a higher number of
vinyl groups per crosslinker molecule. Moreover, the amount
of unreacted pendant vinyl groups in the sols was deter-
mined for a series of reactions with various crosslinkers and
different molar ratios of [Vinylbranch]0/[Initiator]0. The exper-
imentally determined fraction of pendant acrylate groups
(fA,p) was significantly higher than the value theoretically
predicted for gelation at a high targeted crosslink density.
The discrepancy between the theoretical and experimental
FIGURE 5 Comparison of calculated results (solid line) and ex-
perimental results (dots) of the fraction of pendant acrylate
groups (fA,p) as a function of MA conversion during the ATRP
of MA and pentaacrylate crosslinker 5A.
2022
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
12 Oh, J. K.; Drumright, R.; Siegwart, D. J.; Matyjaszewski, K.
fA,p increased with increasing number of vinyl groups per
crosslinker molecule (5A > 4A > 3A). These results help to
better understand the gelation process in controlled radical
copolymerization of monomer and crosslinkers with variable
multiplicities.
Prog Polym Sci 2008, 33, 448–477.
13 Yu, Q.; Zeng, F. Q.; Zhu, S. P. Macromolecules 2001, 34,
1612–1618.
14 Gao, H.; Min, K.; Matyjaszewski, K. Macromolecules 2007,
W. Van Camp thanks the Research Foundation-Flanders (FWO)
for a postdoctoral fellowship and mobility allowance. F. Du Prez
is grateful to the Belgian Program on Interuniversity Attraction
Poles initiated by the Belgian State, Prime Minister’s office
(Program P6/27) for financial support. The financial support
from NSF (DMR-05-49,353) and the CRP Consortium at Carne-
gie Mellon University is also greatly appreciated.
40, 7763–7770.
15 Gao, H.; Li, W.; Matyjaszewski, K. Macromolecules 2008, 41,
2335–2340.
16 Gao, H.; Miasnikova, A.; Matyjaszewski, K. Macromolecules
2008, 41, 7843–7849.
17 Li, W.; Gao, H.; Matyjaszewski, K. Macromolecules 2009, 42,
927–932.
18 Bannister, I.; Billingham, N. C.; Armes, S. P. Soft Matter
REFERENCES AND NOTES
2009, 5, 3495–3504.
1 Landin, D. T.; Macosko, C. W. Macromolecules 1988, 21,
19 Bannister, I.; Billingham, N. C.; Armes, S. P.; Rannard, S. P.;
846–851.
Findlay, P. Macromolecules 2006, 39, 7483–7492.
2 Zhu, S.; Tian, Y.; Hamielec, A. E.; Eaton, D. R. Polymer 1990,
20 Bouhier, M. H.; Cormack, P. A. G.; Graham, S.; Sherring-
31, 154–159.
ton, D. C.
2375–2386.
J Polym Sci Part A: Polym Chem 2007, 45,
3 Kannurpatti, A. R.; Anseth, J. W.; Bowman, C. N. Polymer
1998, 39, 2507–2513.
21 Isaure, F.; Cormack, P. A. G.; Graham, S.; Sherrington, D. C.;
Armes, S. P.; Butun, V. Chem Commun 2004, 1138–1139.
4 Bastide, J.; Leibler, L. Macromolecules 1988, 21, 2647–2649.
22 Li, Y. T.; Armes, S. P. Macromolecules 2005, 38, 8155–8162.
23 Wang, A. R.; Zhu, S. P. Polym Eng Sci 2005, 45, 720–727.
5 Braunecker, W. A.; Matyjaszewski, K. Prog Polym Sci 2007,
32, 93–146.
6 Matyjaszewski, K.; Xia, J. H. Chem Rev 2001, 101, 2921–2990.
24 Wang, A. R.; Zhu, S. P. J Polym Sci Part A: Polym Chem
2005, 43, 5710–5714.
7 Kamigaito, M.; Ando, T.; Sawamoto, M. Chem Rev 2001, 101,
3689–3745.
25 Yu, Q.; Zhang, J. Z.; Cheng, M. L.; Zhu, S. P. Macromol
Chem Phys 2006, 207, 287–294.
8 Tsarevsky, N. V.; Matyjaszewski, K. Chem Rev 2007, 107,
2270–2299.
26 Gao, H. F.; Polanowski, P.; Matyjaszewski, K. Macromole-
cules 2009, 42, 5925–5932.
9 Matyjaszewski, K.; Tsarevsky, N. V. Nat Chem 2009, 1,
276–288.
27 Gao, H. F.; Min, K.; Matyjaszewski, K. Macromolecules 2009,
42, 8039–8043.
10 Gao, H.; Matyjaszewski, K. Prog Polym Sci 2009, 34,
317–350.
28 Gao, H.; Tsarevsky, N. V.; Matyjaszewski, K. Macromole-
11 Ide, N.; Fukuda, T. Macromolecules 1999, 32, 95–99.
cules 2005, 38, 5995–6004.
EFFECT OF CROSSLINKER MULTIPLICITY, VAN CAMP ET AL.
2023