Fast monomers: Factors affecting the inherent reactivity of acrylate monomers in photoinitiated acrylate polymerization

Article abstract of DOI:10.1021/ma021785r

A systematic study on the effect of molecular structure on the photoinitiated polymerization of acrylates was undertaken. Initially, the research was focused on the effect of hydrogen bonding, and it was found that preorganization via hydrogen bonding enhances the maximum rate of polymerization (R_p). This hydrogen bonding facilitated preorganization also affected the tacticity of the resultant polymer. Next, the effect of polarity as represented by the calculated dipole moment (μ_calc) of a given monomer was investigated. A direct linear correlation between R_p and the calculated Boltzmann-averaged dipole moment (μ_calc) was observed. The R_p-μ_calc correlation holds for pure monomers, mixtures of monomers, and even mixtures of monomers with inert solvents. This correlation enables the rational design of monomers with a required reactivity. In addition, these studies suggest that the propagation step of polymerization is influenced by hydrogen bonding while the dipole moment influences the termination rate constant. These two mechanistic explanations can be regarded as complementary factors that influence the speed of acrylate polymerization.

Full text of DOI:10.1021/ma021785r

Macromolecules 2003, 36, 3861-3873
3861
Fast Monomers: Factors Affecting the Inherent Reactivity of Acrylate
Monomers in Photoinitiated Acrylate Polymerization
J oh a n F . G. A. J a n sen ,* Aylvin A. Dia s, Ma r k o Dor sch u , a n d Betty Cou ssen s
DSM Research, P.O. Box 18, 6160 MD Geleen, The Netherlands
Received December 26, 2002; Revised Manuscript Received April 3, 2003
ABSTRACT: A systematic study on the effect of molecular structure on the photoinitiated polymerization
of acrylates was undertaken. Initially, the research was focused on the effect of hydrogen bonding, and
it was found that preorganization via hydrogen bonding enhances the maximum rate of polymerization
(Rp). This hydrogen bonding facilitated preorganization also affected the tacticity of the resultant polymer.
Next, the effect of polarity as represented by the calculated dipole moment (µcalc) of a given monomer was
investigated. A direct linear correlation between Rp and the calculated Boltzmann-averaged dipole moment
(µcalc) was observed. The Rp-µcalc correlation holds for pure monomers, mixtures of monomers, and even
mixtures of monomers with inert solvents. This correlation enables the rational design of monomers with
a required reactivity. In addition, these studies suggest that the propagation step of polymerization is
influenced by hydrogen bonding while the dipole moment influences the termination rate constant. These
two mechanistic explanations can be regarded as complementary factors that influence the speed of
acrylate polymerization.
In tr od u ction
Photoinitiated polymerization of acrylates and meth-
acrylates is one of the most efficient processes for the
rapid production of polymeric cross-linked materials
with defined properties. This very efficiency is the
reason that photoinitiated polymerization is widely
employed in the performance applications where em-
phasis is put on the mechanical as well as optical
properties. These applications are typically dental re-
storative fillers, fiber-optic coatings, optical adhesives,
aspherical lenses for CD applications, and the prepara-
tion of contact lenses.1
Although the use of photoinitiated polymerization is
widespread, fundamental understanding of factors that
affect polymerization speed remains limited. The pre-
ferred method to increase the polymerization rate is to
increase photoinitiator concentration, use of more ef-
ficient photoinitiators, or to increase reactive group
functionality of the oligomers and monomers used.
Regarding mechanistic kinetic investigations, few de-
tailed investigations can be found for acrylates. How-
ever, with methacrylates kinetic investigations2 have
been performed, using calorimetric methods (photodif-
ferential scanning calorimetry (photo-DSC) or spectro-
scopically, e.g., real-time Fourier transform infrared
spectroscopy (RT-FTIR)). Investigations using these
time-resolved techniques have led to several models like
bimolecular termination combined with reaction diffu-
sion-controlled kinetics,3 monomolecular termination in
the glassy region,4 primary radical termination,5 chain
length-dependent termination,6 random walk for het-
erogeneity,7 primary cyclization,8 and the effect of
monomer functionality.9 It is assumed that acrylates
will react according to similar pathways as for meth-
acrylates studied thus far. However, there is enough
evidence in the literature that acrylates behave differ-
ently. Berchthold et al. found a notable difference in the
addition of small amounts of thiols as chain transfer
F igu r e 1. RT-FTIR time conversion profiles of tetrahydro-
furfuryl acrylate 36 and glycerolcarbonate acrylate 57 under
nitrogen initiated with 1% w/w hydroxy cyclohexyl phenyl
ketone.
agents, which have a large impact on the polymerization
rate of methacrylates but almost no influence on the
polymerization rate of the corresponding acrylates.10
Comprehension of the influence of molecular structure
on acrylate reactivity has been sought ever since the
first publications with acrylates and methacrylates that
reported high inherent reactivities. Figure 1 illustrates
the differences in polymerization behavior observed with
RT-FTIR of acrylates with different molecular struc-
tures.
Deckers11 report on acrylates with a very high intrin-
sic reactivity assigned importance to hydrogen abstrac-
tion as a potential reason for high Rp. Andrzejewska12
reported a heteroatom effect in the side chain which led
to higher reactivity. Enhanced polymerization rates
were also found by performing the photoinitiated po-
lymerization in ordered liquid crystalline phases as was
elegantly demonstrated by Guymon.13 These results
stimulated our effort to understand, on a molecular
level, the features that contribute to the high intrinsic
polymerization reactivity of acrylates. On the basis of
such a fundamental understanding, the rational design
of acrylates and blends thereof with required designed
* To whom correspondence should be addressed. E-mail:
johan-fga.jansen@dsm.com.
10.1021/ma021785r CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/09/2003

3862 J ansen et al.
Macromolecules, Vol. 36, No. 11, 2003
Fusion UV curing conveyor system equipped with F600 D 6
kW bulbs. The samples were given the dose required to reach
>97% conversion (determined by NMR). Employing higher
doses resulted in the appearance of unwanted and yet un-
identified signals, which may be attributed to photoinitiator
and polymer/oligomer based secondary hydrogen abstraction
reactions. In fact, even with lauryl methacrylate, insoluble gel
particles were obtained upon using a high UV dose.17 The
tacticity was determined by 13C NMR using a Bruker ARX 400
in DMSO-d6 at 80 °C. A Bruker AMX 200 was used for the
standard 1H and 13C NMR measurements employing DMSO-
d6 as solvent.
Ca lcu la tion of Dip ole Mom en ts. Boltzmann-averaged
dipole moments (µcalc) were calculated in the following way.
First, for a given acrylate a set of starting configurations that
considered all possible bond rotations were generated. This
was done by means of the Discover 95 program.18a Torsional
angles were considered to depend on the type of bond; e.g., for
a bond between two sp3 carbons the angles taken into account
are those corresponding to the two possible gauche conforma-
tions and the trans conformation. The number of configura-
tions generated is thus dependent on both the number of bonds
and their type. For example, for five sp3-like bonds one has 35
) 243 configurations. As a consequence, for some of the
acrylates, the total number of configurations was a few
thousand. All these configurations were then minimized at the
AM1 level using MOPAC 6.0 with the convergence criterion
for the maximum gradient (GNORM) set to 0.05. The resulting
structures were sorted by energy, and only the unique struc-
tures having a heat of formation differing less than 3 kcal/
mol from the heat of formation of the global minimum
structure were retained. Whether or not structures were
unique was determined by comparing their heats of formation
and their dipole moments in the following way. First, struc-
tures were considered to be identical if their heats of formation
differ at most 0.01 kcal/mol. Nevertheless, structures that were
considered to be identical on the basis of this energy criterion
were considered to be unique if their dipole moments differed
more than 0.2 D. Having determined the unique structures,
the Boltzmann-weighted dipole moment was consequently
evaluated according to the formula
F igu r e 2. Acrylate monomers with a high Rp as reported by
Decker.
reactivities can be envisaged, instead of the trial-and-
error approach applied so far. The pioneering work of
Decker, demonstrating that the molecules depicted in
Figure 2 polymerized with a high Rp,11 served as
starting point for our investigations.
On the basis of these results, we decided to focus our
research on the influence of hydrogen bonding and
polarity (as defined by the calculated dipole moment)
as starting rationales for the high polymerization rates
observed in the literature.14 New high reactivity acrylate
monomers are introduced as part of the study.
Exp er im en ta l Section
Most of the monomer preparations are straightforward
organic procedures; hence, only one representative example
is given. Where the synthesis of a monomer is more compli-
cated, the full experimental details are given. All the chemicals
were used as obtained from Aldrich or other sources, except
THF, which was distilled over Na/benzophenone, dichlo-
romethane, which was distilled over P2O5, and pyridine,
triethylamine, and acryloyl chloride, which were freshly
distilled prior to use.
The RT-FTIR setup employed is a transmission reflection
cell with heating and nitrogen inerting capability.15 A 200 W
medium-pressure mercury lamp that delivered a UV light
intensity of 135 mW/cm2 at the sample position as measured
with a Solatel Sola-scope 1 was employed. The sample thick-
ness for nonlaminated samples was 7 µm, unless otherwise
stated. A nitrogen purge was applied for at least 5 min prior
to RT-FTIR measurements, and the temperature was main-
tained at 25.0 ( 1.0 °C. All RT-FTIR measurements were
performed in triplicate, with a spectral sampling rate of 0.038
s per IR spectrum (at 4 cm-1 spectral resolution). In the rare
cases where RT-FTIR experiments were performed on samples
that were laminated, the light intensity at the sample was 45
mW/cm2 and the sample thickness was 50 µm. Differentiation
in order to obtain Rp was performed with an in-house-
developed Fortran program using the following constraints:
maximum error in time resolution 0.005 s, maximum error in
conversion 2%, and the fit should be differentiable twice. For
the kinetic measurements a summation, per nanometer, of the
emission spectrum of the lamp combined with the absorption
spectrum of the photoinitiator was taken for the calculation
of the amount of the photons absorbed. In the kinetic analysis
an optically thin film was assumed, thereby eliminating the
need to know the exact film thickness, i.e., Rp ) (kp/kt1/2)[M]-
(æI0ꢀ[PI])1/2 can be used instead of Rp ) (kp/kt1/2)[M](æIv(1 -
e-∆H /RT
j
µcalc
)
Dj
)
Djpj
∑
∑
e-∆H /RT
j
j
i
∑
i
with Dj the dipole moment of conformation j, ∆Hj the difference
between the heat of formation of conformation j and the heat
of formation of the global minimum conformation, T the
absolute temperature, and R the Boltzmann constant; pj is the
probability of finding the molecule in conformation j at the
temperature T. T is set to 298.15 K. The summation over j
runs over all unique structures. Sorting of structures, retaining
only the unique ones and the calculation of (µcalc) was done by
means of a FORTRAN program developed in-house.
The advantage in considering the Boltzmann-weighted
dipole moment instead of the dipole moment of the global
minimum structure is that the former approach takes into
account the fact that several conformations can be accessible
at T. It is implicit that when the dipole moments of the
accessible conformations are significantly different, the value
of (µcalc) may differ significantly from the dipole moment of the
global minimum. The Boltzmann-weighted dipole moment
(µcalc) therefore provides a much more realistic description of
the system.
Ca lcu la tion of Ra d ica l Ch a r ges. ESP charge calculations
have been performed with the Gaussian98 package18b at the
B3LYP/6-31G* level. In the gas phase a full geometry optimi-
zation was carried out whereas the ESP charges obtained for
D ) 0.9 and D ) 5 were obtained by performing single-point
calculations on the B3LYP/6-31G* gas-phase geometry using
the SCI-PCM method employing the default isodensity value
of 0.0004 au.
10dꢀ[PI] 1/2. The rest of the kinetic analysis was performed
)
assuming steady-state conditions as described by Decker.16
All RT-FTIR measurements were performed employing 1
wt % Irgacure 184 (2-hydroxycyclohexyl phenyl ketone, HCPK,
Ciba) as the photoinitiator. All photoinitiated polymerizations
were performed to an acrylate double bond conversion >96%
(based on IR). The conversion was cross-checked with selected
samples by 1H NMR, 13C NMR, and Raman spectroscopy at
various conversions. The values obtained were in good agree-
ment (within 3% error, although the NMR techniques tend to
become less accurate at high conversions, i.e., above 90%
double bond conversion). Photoinitiated polymerizations of the
methacrylate monomers in order to determine tacticity of the
resultant polymers were performed under nitrogen using a

Macromolecules, Vol. 36, No. 11, 2003
Fast Monomers 3863
Synthesis of Ethyl-O-urethane-N-ethyl Acrylate19 5; Typical
Procedure. To a stirred cooled solution at 0 °C of 14.22 g (107
mmol) of N-2-hydroxyethyl-O-ethylurethane and 8.45 g (107
mmol) of pyridine in 300 mL of THF, purged with dry air, was
slowly added 10.1 g (112 mmol, 1.05 equiv) of acryloyl chloride
while maintaining the temperature below 5 °C. After the
addition was complete, the reaction mixture was allowed to
warm to room temperature and was then stirred at this
temperature for an additional 6 h.
Workup. The pyridinium salt was filtered off, followed by a
quick wash with 20 mL of water, 0.1 N hydrochloric acid,
saturated NaHCO3 solution, and brine (where the separation
was tedious diethyl ether was added). After drying over Na2-
SO4, THF was removed in vacuo, yielding 6.1 g (33 mmol, 31%)
of ethyl-O-urethane-N-ethyl acrylate as crude product with a
purity based on NMR of 97%.
propane carbonate was carefully neutralized with p-toluene-
sulfonic acid over a neutralization period off at least 1 h. This
neutralization step is very critical. In basic conditions the
depolymerization of the oligomeric polycarbonate hardly oc-
curs. However, in an acidic environment a rapid, acid-catalyzed
decarboxylation occurs as well, after which a hydroxy func-
tional oxetane can be isolated.
The polymeric trimethylolpropane carbonate was trans-
ferred into a preheated (150 °C) dropping funnel. The polymer
was dropped slowly (maximum rate 100 mL/h) into a heated
(250 °C) 250 mL round-bottom flask equipped with a distilla-
tion setup and charged with 1 g of tin flakes under a reduced
pressure (8 mbar). Depolymerization occurred immediately,
and the crude product was distilled over. As soon as the
depolymerization was no longer spontaneous, the reaction was
stopped and only 50% crude product was obtained. Unfortu-
nately, the crude product obtained in this way still contained
some (15%) hydroxy functional oxetane. All our attempts to
separate these two alcohols by distillation failed. Often the
amount of oxetane increased during distillation, which might
indicate a temperature-induced decarboxylation reaction. There-
fore, we decided to continue with the crude alcohol and purify
the acrylate. It should be noted that the NMR spectra of the
oxetane and the cyclic carbonate are similar; at first glance,
the weak carbonyl absorption in 13C NMR at 154.8 ppm and
the ethyl CH2 are the distinct differences. Carbonate (6-CCOH)
1H NMR (DMSO-d6): 5.00 (t, J ) 5.2 Hz, 1H), 4.20 (AB, ν )
15.2 Hz, J ) 10.2 Hz, 4H), 3.38 (d, J ) 5.2, 2H), 1.37 (q, J )
7.6 Hz, 2H), 0.82 (t, J ) 7.6 Hz, 3H). 13C NMR (DMSO-d6):
154.5 (s), 76.6 (t), 61.8 (t), 35.0 (s), 21.8 (t), 7.2 (q). IR (neat)
cm-1: 1734. For comparison, the 1H NMR spectrum of the
oxetane: 1H NMR: (DMSO-d6): 5.00 (t, J ) 5.3 Hz, 1H), 4.22
(AB, ν ) 21.4 Hz, J ) 5.6 Hz, 4H), 3.48 (d, J ) 5.3, 2H), 1.62
(q, J ) 7.6 Hz, 2H), 0.83 (t, J ) 7.6 Hz, 3H).
Alternative Workup. The pyridinium salt was filtered off,
followed by removal of the THF at a vacuum of 10 mbar,
yielding 18.2 g (97 mmol, 90%) of crude ethyl-O-urethane-N-
ethyl acrylate with a purity based on NMR of 85%.
Distillation Procedure. 1 g of the crude product was used to
determine the boiling point of the acrylate by slowly raising
the temperature at a vacuum of 2 mbar until all remaining
THF, pyridine, and acryloyl chloride were distilled. At 135 °C
the boiling temperature remained constant for approximately
10 min before the acrylate started to polymerize in the
distillation flask due to the prolonged heating. Next, 15 g of
the remaining crude product was distilled by quickly raising
the temperature to 150 °C, taking a large first run, followed
by the product. As soon as any polymerization was noticed or
suspected (fumes), the distillation was stopped, yielding 9 g
of product, which still contained some trace impurities pre-
sumably due to the stabilizer used in the acryloyl chloride. To
obtain pure product, i.e., no other signals detected on GC, the
distillation had to be repeated three times after which only 2
g (10% yield) of pure ethyl-O-urethane-N-ethyl acrylate was
6-Ring Carbonate Acrylate 58. 100 g (0.63 mol) of 6CCOH
was dissolved in 1 L of dry CH2Cl2 at room temperature, and
70 g (0.7 mol) of triethylamine was added, after which the
reaction mixture was cooled to 0 °C. 60 g (0.66 mol) of acryloyl
chloride was dissolved in 1 L of CH2Cl2, and this solution was
added dropwise to the reaction mixture while the temperature
was maintained between 0 and 5 °C. After the addition was
complete the reaction mixture was stirred for an additional 2
h between 0 and 5 °C before the reaction mixture was allowed
to warm to room temperature overnight. Filtration followed
by evaporation of the solvents yielded 140 g of crude product.
The distillation procedure described above was followed and
after four distillations (195 °C, 0.1 mbar! so a severe risk of
acrylate polymerization during the distillation exits); 10 g (46
mmol, 8%) of pure 6-ring carbonate acrylate 58 acrylate was
1
obtained. H NMR (CDCl3, 200 MHz): δ (ppm) 6.44 (dd; 1H;
J ) 17.22; J ) 1.48), 6.15 (dd; 1H; J ) 10.33; J ) 17.22), 5.88
(dd; 1H; J ) 10.33; J ) 1.48), 5.72 (s br; 1H; NH), 4.23 (m;
4H), 3.21 (dt; 2H; J ) 6.77), 1.14 (t; 3H; J ) 6.85). 13C NMR
(CDCl3, 50 MHz): δ (ppm) 166.7 (s), 157.0 (s), 132.0 (t), 128.8
(d), 63.6 (t), 63.0 (t), 36.5 (t), 15.8 (q). IR (neat): abs cm-1
,
3345 (N-H), 1728 (CdO), 1534 (N-H def), 810 (CdC-H).
All liquid compounds were distilled at least three times
before they were obtained in a pure form, except the ureas,
which were recrystallized three times. Where bulb-to-bulb
distillation was used approximately five distillations are
required; however, the overall yields are higher due to the
lower temperatures employed (for ethyl-O-urethane-N-ethyl
acrylate; 23% overall yield). Another possible distillation
procedure from which stabilizer-free acrylate can be obtained
is distilling carefully over powdered KOH. However, this
procedure is only applicable where the acrylate side group is
stable to base. In our hands, neither inhibitor removing
columns nor chromatography yielded inhibitor-free acrylates.
Synthesis of 6-Ring Cyclic Carbonate Acrylate 58. Synthesis
of Trimethylolpropane Carbonate (6-CCOH). A 3 L round-
bottom three-neck flask, equipped with a mechanical stirrer,
distillation setup, and a small dropping funnel, was charged
with 1000 g of trimethylolpropane and 920 g of diethyl
carbonate. After heating the reaction mixture to 110 °C, 0.841
g of KOH dissolved in 2 mL of water was added dropwise, after
which the temperature was raised to 120 °C. The reaction
mixture was stirred at 120 °C for 12 h, during which ap-
proximately 400 g of ethanol was distilled over. After distilling
off the remaining ethanol at a reduced pressure (800 mbar),
the composition of the reaction mixture was checked by GC.
This revealed that still 8% trimethylolpropane remained in
the reaction mixture. Therefore, 100 g of diethyl carbonate and
0.4 g of KOH dissolved in water was added, and the reaction
was continued at 120 °C for an additional 8 h, after which the
ethanol and excess diethyl carbonate were stripped off and the
conversion checked by GC. This procedure had to be repeated
two times before full conversion (based on GC) of trimethylol-
propane was reached. The remaining polymeric trimethylol-
1
obtained. H NMR (CDCl3, 200 MHz): δ (ppm) 6.44 (dd; 1H;
J ) 17.22; J ) 1.48), 6.15 (dd; 1H; J ) 10.33; J ) 17.22), 5.88
(dd; 1H; J ) 10.33; J ) 1.48), 4.3-4.1 (m, 6H) 1.40, q, J ) 7.6
Hz, 2H) 0.85 (t, J ) 7.6 Hz, 3H). 13C NMR (CDCl3, 50 MHz):
δ (ppm) 166.5 (s), 148.3 (s), 132.0 (t), 128.8 (d), 72.5 (t), 62.9
(t), 34.6 (s), 23.3(t), 7.3 (q). IR (neat): abs cm-1, 1730 (CdO
broad) 810 (CdC-H). After all the difficulties with the
synthesis of this alcohol we contacted Bayer.20 They confirmed
that the synthesis is not a trivial matter and kindly sent us a
sample of this 6-ring carbonate alcohol.
Synthesis of Carbonate Urethane Acrylate 65. Synthesis21
of Glycerol Carbonate 62: Procedure 1. A round-bottom
flask charged with 725 g (7.7 mol) of glycerol 60, 500 g of
ethylene carbonate, and 0.4 g of NaOH was stirred at 140 °C
for 2 h. Neutralization with p-toluenesulfonic acid to pH ) 6
was followed by distilling off the ethylene glycol. Next, 250 g
of ethylene carbonate and 0.4 g of NaOH were added, and
the reaction mixture was stirred again at 140 °C for 2 h.
This procedure was repeated twice, after which the neutraliza-
tion was performed to pH ) 5. At this acidity the distillation
of the product can be performed at 160 °C at 0.01 mbar,
yielding 650 g (5.5 mol, 70%) of glycerol carbonate with a
purity >97% (pure by NMR). On the basis of GC, the
impurities that remained were mainly glycerol, glycidol, and
some oligomeric material, which can be removed by two
distillations.

3864 J ansen et al.
Macromolecules, Vol. 36, No. 11, 2003
Procedure 2. A round-bottom flask was charged with 45 g
(0.34 mol) of 3-allyl-1,2-propanediol 61, 250 mL of CH2Cl2, and
130 mL of triethylamine (0.9 mol), and the reaction mixture
was cooled to 0 °C. 200 mL of a solution of phosgene in toluene
(20% in toluene, 0.4 mol) was added dropwise at such a rate
that the reaction temperature was kept below 10 °C. After the
addition the mixture was stirred for a further 2 h at 0 °C before
it was allowed to warm to room temperature overnight.
Washing with water, 5% sodium bicarbonate solution, and
water, drying over Na2SO4, and evaporation of the solvent
followed by bulb-to-bulb distillations yielded 36 g (0.22 mol,
67%) of the allyl protected cyclic carbonate. 24.5 g (0.16 mol)
of the allyl protected cyclic carbonate was dissolved in 300 mL
of methanol. After stopping the stirrer, nitrogen was flushed
for 10 min followed by the addition of 3.7 g of Pd/C (10% Pd)
and subsequently 2.5 g of p-toluenesulfonic acid. The reaction
mixture was then stirred at room temperature, and the
conversion was monitored by 1H NMR. After 6 days the
deprotection was complete. The reaction mixture was filtered
over neutral Al2O3. The Al2O3 was subsequently rinsed with
500 mL of methanol. The methanol fractions were combined
and concentrated. Distillation (0.01 mbar, 160-170 °C) yielded
14 g (0.13 mol, 80%) of pure glycerol carbonate 62 (purity
>99%, GC). 1H NMR (CDCl3, 200 MHz): δ (ppm) 5.4 (s br;
1H), 4.9-4.8 (m; 1H), 4.6-4.4 (m; 2H), 3.9-3.5 (m, 2H). 13C
NMR (CDCl3, 50 MHz): δ (ppm) 155.5 (s), 77.1 (d), 66.0 (t)
60.6 (t). IR (neat): abs cm-1, 1734 (CdO broad).
Sch em e 1. Syn th esis of Mon om er s Ca p a ble of
Hyd r ogen Bon d in g a n d Th eir Non -Hyd r ogen -Bon d in g
Con tr ol Mon om er s
Synthesis of the Glycerol Carbonate Active Ester 63. A round-
bottom flask was charged with 38.23 g (0.322 mol) of glycerol
carbonate 62 and 28 g (0.35 mol) of pyridine and cooled to 0
°C before 50.35 g (0.322 mol) of phenyl chloroformate was
added dropwise at such a rate that the reaction temperature
was kept below 5 °C. After the addition was complete the
reaction mixture was kept at this temperature overnight. The
salts formed were filtered off, and the remaining CH2Cl2 layer
was washed with water, dilute hydrochloric acid (0.01 N),
water, a sodium bicarbonate solution, water, and brine fol-
lowed by drying over Na2SO4, and evaporation of the solvent
yielded 66.8 g of the active ester 63 (0.28 mol, 87%), which
was pure enough to be employed in the subsequent reaction.
1H NMR (DMSO, 200 MHz): δ (ppm) 7.7-7.4 (m; 5H), 5.2-
5.1 (m; 1H), 4.7-4.3 (m; 4H).
Synthesis of Hydroxyurethane Glycerol Carbonate 64. 17.55
g (74 mmol) of active ester 63 was dissolved in 200 mL of THF
and cooled to -5 °C, and 4.5 g (74 mmol) of ethanolamine in
200 mL of dry THF was added dropwise, while maintaining
the temperature below 0 °C. (If performed slowly, no exotherm
should be noticeable.) After the addition is complete (ap-
proximately 3 h) the reaction mixture was stirred for 3 days
at 0 °C before the temperature was slowly raised to room
temperature. The THF layer is diluted with ether and ex-
tracted with 3 × 100 mL of distilled water. The combined water
layers were washed with CH2Cl2, after which the water is
evaporated in vacuo, yielding 13.6 g (67 mmol, 90%) of hydroxy
urethane glycerol carbonate 64. 1H NMR (DMSO-d6, 200
MHz): δ (ppm) 7.5 (s br; 1H), 5.3 (s br; 1H), 4.9-4.8 (m; 1H),
4.7-4.4 (m; 4H), 3.9-3.5 (m, 4H). 13C NMR (DMSO-d6, 50
MHz): δ (ppm) 161.7 (s), 155.9 (s), 75.8 (d), 65.7 (t) 63.8 (t),
62.7 (t), 45.5 (t). IR (neat): abs cm-1, 3400-3100 (NH, OH),
1730 (CdO broad). It is difficult to remove the last traces of
water; therefore, a larger excess acryoyl chloride is used in
the next step.
65 (3.5 mmol, 15%). 1H NMR (DMSO, 200 MHz): δ (ppm)
7.5 (t; 1H; J ) 5.7 Hz), 6.44 (dd; 1H; J ) 17.22; J ) 1.48),
6.15 (dd; 1H; J ) 10.33; J ) 17.22), 5.88 (dd; 1H; J ) 10.33;
J ) 1.48), 5.0 (m; 1H), 4.6-4.0 (m; 6H), 3.3-3.1 (m, 2H).
13C NMR (DMSO, 50 MHz): δ (ppm) 167.0 (s), 161.7 (s), 155.9
(s), 131.7 (t), 128.4 (d), 74.8 (d), 65.9 (t), 63.4 (t), 62.9 (t), 45.4
(t). IR (neat): abs cm-1, 3200 (NH), 1730 (CdO broad), 810
(CdC-H).
Resu lts a n d Discu ssion
Mon om er s Ca p a ble of Hyd r ogen Bon d in g. The
high reactivity of the urethane ethyl acrylates depicted
in Figure 2 suggested that the high Rp of these mono-
mers could be attributed to hydrogen bonding. To
evaluate this hypothesis, we performed initial curing
experiments with undecyl amide N-ethyl acrylate, a
monomer capable of hydrogen bonding, and pentyl
amide N-methyl N-ethyl acrylate as a non-hydrogen-
bonding, control monomer with the same secondary
functionality. The Rp of these acrylate monomers were
determined by RT-FTIR. The Rp of the hydrogen-
bonding monomer, undecyl amide N-ethyl acrylate, was
9 mol/(L s) while the non-hydrogen-bonding analogue
pentyl amide N-methyl N-ethyl acrylate gave a much
lower Rp of 1.9 mol/(L s). These results are indicative of
the possible importance of hydrogen bonding on the rate
of acrylate photoinitiated polymerization. These encour-
aging initial results triggered us to systematically vary
the nature of the secondary group of the acrylate. As
monomers capable of hydrogen-bonding amide, ure-
thane, “inverted urethane”, and urea acrylates were
prepared. Analogous ester and carbonate acrylates were
synthesized and evaluated as control compounds since
as they possess no capability for hydrogen bonding. The
synthesis of these compounds is shown in Scheme 1.
The preparation of amide ethyl acrylates via the
oxazoline precursor was rendered difficult by purifica-
tion of the end product. This purification is difficult
because approximately 1% of the corresponding alkyl
ester ethyl acrylamide is formed during synthesis, and
selective distillation of the amide ethyl acrylate was very
difficult. Since we were unable to remove this byproduct
by chromatography or distillation, we abandoned this
Synthesis of Glycerol Carbonate Urethane Acrylate 65. To a
stirred cooled solution at 0 °C, purged with dry air, of 5.2 g
(25 mmol) of hydroxy urethane glycerol carbonate 64 and 8 g
(100 mmol) of pyridine in 300 mL of THF was slowly added 7
g (76 mmol, 3 equiv) of acryloyl chloride in 100 mL of THF
while maintaining the temperature below 5 °C. After the
addition was complete, the reaction mixture was allowed to
warm slowly to room temperature and was stirred at room
temperature for an additional 6 h. After the pyridinium salt
was filtered off, the mixture was cooled to 0 °C before the
mixture was filtered again. The excesses employed were
evaporated off (100 °C, 0.7 mbar), followed by several distil-
lations yielding 0.9 g of glycerol carbonate urethane acrylate

Macromolecules, Vol. 36, No. 11, 2003
Fast Monomers 3865
Ta ble 1. Rp (m ol/(L s)) of P h otoin itia ted P olym er iza tion
of Mon om er s Ca p a ble of Hyd r ogen Bon d in g a n d th e
Non -Hyd r ogen -Bon d in g Con tr ol Com p ou n d s Deter m in ed
by RT-F TIR w ith a Ligh t In ten sity of 135 m W/cm 2
(Med iu m -P r essu r e Mer cu r y La m p )
F igu r e 3. Effect of varying the bridge length between acrylate
and amide group on the maximum rate of photoinitiated
acrylate polymerization.
case of a noncovalent connection no effect on Rp is
expected as the amount of “bifunctional” material is
independend of the bridge but only depended on the
association strength.
Contrary to normal polymerizations, which exhibit
Arrhenius type behavior, increasing the temperature
should lead to a deviation from this behavior since at
elevated temperatures the extent of hydrogen bonding
is reduced. So according to both explanations, lower Rp’s
should be found at elevated temperatures.
Preorganization might influence the tacticity of the
polymer. Therefore, the tacticity of the resultant poly-
mer could be different when compared with polyacry-
lates prepared from monomers, incapable of hydrogen
bonding. However, in the case of a noncovalent connec-
tion again no effect is expected, in this case the acrylate
monomers would not be aligned and the normal tacticity
should be found.
Changes in tacticity would also discriminate between
inter- and intramolecular hydrogen bonding, thereby
adding further validity to the preorganization concept.
Only with preorganization via intermolecular hydrogen
bonding can increased isotacticity be expected. In the
case of intramolecular hydrogen bonding it is still
possible that Rp enhancements could be found due to a
charging of the acrylate double bond; however, no effect
on the tacticity would be expected. As another alterna-
tive explanation for the enhanced Rp’s, the higher
viscosity of the hydrogen-bonding monomers could be
proposed. However, changes in tacticity cannot be
explained by this theory.
Effect of Br id ge Len gth . As mentioned earlier, one
consequence of preorganization is that when the hydro-
gen-bonding moiety is at a larger distance from the
reactive acrylate group, there will be a critical distance
beyond which the preorganization due to the hydrogen
bonding will have little or no effect on Rp. This is
because conformational mobility of the skeletal bonds
between the hydrogen-bonding moiety, and the acrylate
double bond will reach a level whereby the acrylate
double bond can be regarded as isolated from the
hydrogen-bonding moiety. Verification of this prediction
based on the preorganization theory was done by
preparing and evaluating several ethyl amide N-alkyl
acrylates, in which the alkyl chain was varied in length
from ethyl to hexyl (compounds 1, 25, 26, 27, 28). The
results of these RT-FTIR experiments are shown in
Figure 3.
otherwise versatile synthetic route. To obtain the com-
pounds with high purity (>99% based on GC), all the
syntheses were performed via the acid chloride, chlo-
roformate, or isocyanate starting materials. All the
syntheses gave the crude compounds in yields between
70 and 90%. Most of the alkyl urea ethyl acrylates
(except ethyl urea ethyl acrylate) were solids at room
temperature. As a consequence, the Rp of the ureas were
determined at 50 °C when all the samples were in their
liquid state. In Table 1 the Rp of amide, urethane,
“inverted urethane”, and urea acrylates as well as their
non-hydrogen-bonding analogues are shown.
It is evident from Table 1 that monomers capable of
forming hydrogen bonds generally exhibit higher Rp’s
compared to their non-hydrogen-bonding analogues.
Comparing the ethyl-substituted compounds around
3-6 times higher Rp’s are observed. These findings
demonstrate that hydrogen bonding has a pronounced
effect on the maximum rate of these photoinitiated
acrylate polymerizations.
How can these effects be rationalized? One explana-
tion is that via hydrogen bonding the monomers are
dynamically noncovalently connected, i.e., behave partly
like a difunctional species. Such a scenario is known
for methacrylic acid,22 and in that case it has been
established that the propagation rate constant (kp) was
hardly influenced but that the termination rate constant
(kt) has been dramatically reduced.
Another plausible explanation for these enhanced Rp’s
is that hydrogen bonding facilitates preorganization in
these acrylates, thereby forcing the acrylate double
bonds in close proximity to each other. As a conse-
quence, the propagation reaction rate constant (kp) will
be enhanced resulting in higher Rp. Organization of
monomers via hydrogen bonding by either explanation
has several consequences.
There should be a critical distance between the
acrylate group and the hydrogen-bonding moiety (alkyl
bridge length) beyond which there is no effect on Rp in
the case of the preorganization theory, whereas in the
The Rp’s of the ethyl amide N-alkyl acrylates are
gradually reduced going from ethyl to pentyl, in line

3866 J ansen et al.
Macromolecules, Vol. 36, No. 11, 2003
Sch em e 2. P r op osed Mech a n ism by Wh ich
P r eor ga n iza tion via Hyd r ogen Bon d in g Cou ld Lea d
to th e F or m a tion of Isota ctic P olym er
F igu r e 4. Dependency of the Rp of undecyl amide N-ethyl
acrylate on the reaction temperature.
Ta ble 2. Ta cticity of P olym eth a cr yla tes, Ba sed on 13C
NMR, F or m ed by P h otoin itia ted P olym er iza tion
with the theory that the effect of the hydrogen bonding
will be reduced at a larger distance from the acrylate
group. Furthermore, ethyl amide N-hexyl acrylate ex-
hibits a Rp similar to monomers not capable of forming
hydrogen bonds like the ester and carbonate functional
acrylates in Table 1. The occurrence of such a critical
distance encourages our belief in the preorganization
of acrylate monomers via hydrogen bonding instead of
the noncovalent connection theory.
undecyl
pentyl amide
tacticity
(%)
amide N-ethyl-
methacrylate (29) methacrylate (30) methacrylate
N-methyl N-ethyl-
lauryl
syndio
58
28
14
65
33
2
65
31
4
a
iso
anti-Arrhenius behavior cannot discriminate between
the two different theories.
Effect of Tem p er a tu r e. It is recognized that when
reactions are performed at higher temperatures, higher
rates will be observed this is the so-called Arrhenius
behavior. However, if the high Rp is due to preorgani-
zation via hydrogen bonding or via a noncovalent
connection, at higher temperatures where the extent of
hydrogen bonding is reduced, one can expect lower Rp.
This reasoning was put to the test with RT-FTIR
experiments performed at elevated temperatures. To
avoid problems associated with thin films and monomer
evaporation at elevated temperatures, we used the high
boiling undecyl amide N-ethyl acrylate as a monomer
capable of hydrogen bonding, and as a control we
employed lauryl acrylate as monomer not capable of
hydrogen bonding. The results of the experiments with
undecyl amide N-ethyl acrylate are shown in Figure 4.
The temperature-dependent behavior of lauryl acry-
late, which corresponded to an activation energy around
4-5 kJ /mol, is in good agreement with the values
obtained using photo-DSC.2a,c For diacrylates, using RT-
IR up to 90 °C a similar activation energy has been
found.23 Therefore, the behavior of undecyl amide N-
ethyl acrylate clearly deviates from the Arrhenius
behavior of normal acrylates.
Since we are employing RT-FTIR, we are able to
monitor simultaneously the amide shifts as well as the
acrylate conversions. The amide II bond shifts from 1546
cm-1 at 30 °C to 1539 cm-1 at 90 °C, and the N-H shifts
from 3312 cm-1 at 30 °C to 3325 cm-1 at 90 °C. Both
shifts are indicative of reduced hydrogen bonding at
elevated temperatures. Unfortunately, the quantitative
extent of hydrogen bonding at 30 or 90 °C is unknown;
therefore, it is difficult to proportionate the extent of
hydrogen bonding with Rp. It should be noted that with
urethane and urea acrylates we were only able to
observe a clear IR spectral shift toward non-hydrogen
bonding at higher temperatures with the corresponding
N-H absorptions.
Effect on Ta cticity. The effect of preorganization
can be examined further by looking at the tacticity of
the resultant polymer. Only in case where intermolecu-
lar preorganization via hydrogen bonding is significant
can an effect on the resultant polymer tacticity can be
expected. Generally, polyacrylates produced by a free
radical mechanism yield predominantly syndiotactic
material. However if we assume that the polymerization
proceeds via an organized structure, then isotactic
polyacrylate can be envisaged as shown in Scheme 2.
However, it is unlikely that all the monomers will be
fully preorganized in order to give pure isotactic poly-
mer. Experiments were performed in order to see if
there was enhanced isotacticity in the resultant poly-
mers. Since the tacticity can be better determined with
methacrylates instead of acrylates, the experiments to
determine tacticity were performed with the following
methacrylates: undecyl amide N-ethyl methacrylate 29
as monomer capable of hydrogen bonding, pentyl amide
N-methyl N-ethyl methacrylate 30 as control compound
possessing the amide functionality but incapable of
hydrogen bonding, and lauryl methacrylate as reference
sample. The results are shown in Table 2.
The tacticity of poly(lauryl methacrylate) (polyLMA)
produced via photoinitiated polymerization is identical
to polyLMA prepared with thermal initiators like AIBN,
thereby eliminating a possible effect due to the initiator
employed.24 The tacticity of poly(pentyl amide N-methyl
N-ethyl methacrylate) is the same as for polyLMA
(within the errors of the measurements). This strongly
indicates that an amide function itself does not alter
the tacticity of the polymer. Analyzing the poly(undecyl
amide N-ethyl methacrylate) formed we observed an
increased amount of isotactic triads as determined by
NMR. This enhancement is significant and beyond the
instrumental and experimental errors. The enhanced
isotacticity observed with hydrogen-bonding acrylates
suggests that preorganization via intermolecular hy-
drogen bonding plays a dominant role in how the
acrylates groups are oriented toward each other. There-
fore, this provides additional evidence for the hydrogen-
This reduction in Rp, at elevated temperatures, com-
bined with the observed IR shifts, provides additional
evidence validating the fact that hydrogen bonding is a
key issue in these rate enhancements. However, this

Macromolecules, Vol. 36, No. 11, 2003
Fast Monomers 3867
bonding-assisted preorganization, its effects on the Rp,
and the tacticity of resultant polymer.
Although all the data presented above suggest that
hydrogen bonding affects the propagation rate constant
and not the termination constant, the ultimate proof
obviously is the determination of the kinetic constants.
It is difficult to obtain a correct comparison, a difference
in hydrogen bonding but no difference in dipole moment
(see below) for the determination of the rate constants,
although the first results indicate that indeed kp is
affected and not kt. This item will be discussed further
when the effects of polarity and hydrogen bonding are
combined.
F igu r e 5. Correlation between the calculated Boltzmann-
averaged dipole moment and the maximum rate of photoini-
tiated polymerization as determined with RT-FTIR. Closed
circles, neat monomers; open circles, mixtures of monomers.
Mon om er s Not Ca p a ble of Hyd r ogen Bon d in g.
As mentioned in the Introduction (Figure 2), other
monomers with enhanced reactivity exist for which the
hydrogen-bonding theory offers no explanation since
these monomers are not capable of forming hydrogen
bonds. Therefore, another mechanism that explains the
high reactivity of non-hydrogen-bonding monomers has
to be considered. Decker11 and Bowman25 both mention
hydrogen abstraction reactions, via which cross-links
can be introduced, as the explanation for the high
reactivity. The cross-links could lead to a viscosity
increase,26 earlier start of gelation, thereby allowing an
earlier onset of autoacceleration (Trommsdorf effect27),
which would account for high Rp’s. However, experi-
mental evidence, in the literature, is inconclusive. The
fact that the polymerization of glycerol carbonate acry-
late 57 can be initiated with benzophenone, a Norrish
type II photoinitiator, suggests that under these condi-
tions some of the hydrogens are relatively labile.
However, it does not indicate that these hydrogens are
abstracted during the radical polymerization. It only
demonstrates that they can be abstracted during the
initiation (when nothing else can happen). The other
feature, which is assumed to prove the cross-link
hypothesis, is the formation of gel particles. Again, this
only proves that some of the hydrogen atoms can be
abstracted. In fact, polymerizing lauryl acrylate with a
very high UV dose will result in the formation of gel
particles. Moreover, butyl acrylate polymerized at 60 °C
will also give some gel particles.17 The experimental
data, which suggest that hydrogen abstraction is not
the mechanistic pathway, is not discussed in these
papers. The fact that the oxazolidone functional acrylate
56 has a lower Rp compared to that of glycerol carbonate
acrylate 57 is ignored in the discussions. A consequence
of the hydrogen abstraction theory would be that the
oxazolidone functional acrylate 56, possessing more and
more labile hydrogens (adjacent to a nitrogen atom
instead of an oxygen atom), should react faster than the
glycerol carbonate acrylate 57. On the contrary, the
oxazolidone functional acrylate 56 has a lower Rp
compared to that of the glycerol carbonate functional
acrylate 57. In light of our results and often contradic-
tory reports on the effect of hydrogen abstraction, we
decided to explore whether dipolar interactions could
account for the high intrinsic polymerization reactivity
of certain acrylates, especially since it is well-known
that polarity is important in radical addition reac-
tions.28,29 Tedder29 postulated a second law of free
radical reactions based on polarity effects.
ments nor the dielectric constants are known. Although
the calculated dipole moment of an energy-minimized
structure could give an indicative value, it has to be
realized that in a (neat) solution more conformations
than only the minimum-energy structure are present
and that the average dipole moment could therefore
differ significantly from the dipole of the energy-
minimized structure. Acknowledging this, we calculated
the dipole moments as the Boltzmann-averaged dipole
moments (µcalc), and we believe that the Boltzmann-
averaged dipole moment is a good representation of
dipole of a neat solution. As far as we are aware, this is
the first time that dipole moments are calculated as
Boltzmann-averaged dipole moments. Normally, the
dipole moment of the global energy-minimized structure
is used (µgm). The following series is illustrative for the
large differences which can be found employing these
two methods: compound 36: µcalc ) 2.05 D, µgm ) 1.74
D; compound 38: µcalc ) 2.17 D, µgm ) 2.28 D; compound
48: µcalc ) 2.65 D, µgm ) 3.50 D; compound 49: µcalc
2.93 D, µgm ) 3.2 D; compound 56: µcalc ) 5.3 D, µgm
)
)
4.9 D; compound 57: µcalc ) 5.9 D, µgm ) 6.6 D. The
differences employing monomers not capable of hydro-
gen bonding are still minor compared to monomers
capable of hydrogen bonding: compound 1: µcalc ) 4.31
D, µgm ) 2.83 D; compound 5: µcalc ) 3.24 D, µgm ) 0.93
D; compound 9: µcalc ) 3.23 D, µgm ) 2.11 D; compound
13: µcalc ) 4.64 D, µgm ) 2.6 D; compound 21: µcalc
)
3.3 D, µgm ) 1.83 D. Especially the fact that compounds
5 and 9 possess almost the same Rp as well as the same
(µcalc) but a very different µgm corroborates our believe
that calculating Boltzmann-averaged dipole moments
is the correct procedure.
After some initial experiments in which we applied
the calculated dipole moments of the global energy-
minimized structure, we realized that there could be a
direct correlation between the dipole moment and the
Rp and that the observed deviations might be due to the
incorrect method of calculating the dipole moment. Only
after this realization we started the calculations of the
Boltzmann-averaged dipole moments, instead of the
global energy-minimized dipole moments, for a large set
of monomers and compared them to the Rp values. The
results of this study are shown in Table 3 and Figure 5
(black circles). The results clearly show that for mono-
mers with calculated dipole moments (µcalc) above 3.5
D there is a direct correlation between the Rp and (µcalc).
This correlation between a monomer molecular struc-
ture, its associated polarity, and Rp for acrylate polym-
erization enables the rational design of new monomers
The main problem faced in any attempt to discuss or
experimentally determine acrylate reactivity in terms
of dipolar interactions is the fact that for most of the
interesting reactive monomers neither the dipole mo-

3868 J ansen et al.
Macromolecules, Vol. 36, No. 11, 2003
Ta ble 3. Ma xim u m Ra tes (m ol/(L s)) of P h otoin itia ted P olym er iza tion of Mon om er s Not Ca p a ble of Ca p a ble of Hyd r ogen
Bon d in g As Mea su r ed by RT-F TIR Usin g a Tr a n sflection Cell a t Room Tem p er a tu r e Em p loyin g a Ligh t In ten sity of 135
m W/cm 2 (Med iu m -P r essu r e Mer cu r y La m p ) a n d Th eir Ca lcu la ted Boltzm a n n -Aver a ged Dip ole Mom en ts
with a designed reactivity instead of the trial-and-error
strategies applied so far.
Thus, a direct correlation with (µcalc) and Rp has been
found. So the averaged dipole moment is determining

Macromolecules, Vol. 36, No. 11, 2003
Fast Monomers 3869
F igu r e 7. Illustration of the propagating radical envisaging
the solvent cage and the enhanced charging of the propagating
radical.
F igu r e 6. Correlation between the calculated Boltzmann-
averaged dipole moment and the maximum rate of photoini-
tiated polymerization of tetrahydrofurfuryl acrylate diluted
with dimethyl carbonate (DC) or propylene carbonate (PC) as
determined with RT-FTIR.
dipole moment (µcalc) and the maximum rate of polym-
erization (Rp) for monomers not capable of hydrogen
bonding, and this offers a tool for the rational design of
inherently reactive monomers.
the maximum rate of polymerization. These results open
up two possibilities. The first is that under identical
conditions mixtures of acrylates should obey the same
correlation. The second is that in case small amounts
of highly polar inert solvents are added the rates of
polymerization should increase.
Mech a n ism ? Several possible explanations for this
relationship can be envisaged: (a) The photoinitiator
can be more effective in a more polar medium.32 As more
initiating radicals are generated in this way a higher
rate of polymerization will be found. (b) The activation
energy of propagation can be reduced in a more polar
medium, leading to a faster propagation reaction. (c)
Hydrogen abstraction reactions leading to cross-link
points are facilitated; therefore, an earlier onset of
autoacceleration may occur. (d) A stronger solvent cage
around the propagating radical is formed in a higher
polarity medium, resulting in a lower termination
reaction rate and consequently faster polymerization.33
(e) In a higher dipole medium, the propagating radical
is charged to a greater extent, resulting, due to the
increased electrostatic repulsion between the terminat-
ing radicals, in a reduction of the termination reaction
rate.34
To evaluate the first possibility, we examined the
photoinitiated polymerization of various mixtures of
tetrahydrofurfuryl acrylate 36 (µcalc ) 2.05 D) and
oxazolidone-N-ethyl acrylate 56 (µcalc ) 5.3 D). The
results were in line with the µcalc-Rp correlation and
are shown in Figure 5 (open circles). These findings
demonstrate that the µcalc-Rp correlation is valid for
mixtures of monomers as well as neat monomers.
The second possibility takes this theory one step
further as it suggests that rate enhancements can be
achieved by the addition of polar inert solvents, which
is contradictory to the general belief that dilution with
an inert solvent automatically will lead to a reduction
of the polymerization rate (due to the reduced monomer
concentration). To examine the role of the solvent, we
diluted tetrahydrofurfuryl acrylate 36 with a high polar
(propylene carbonate, D ) 5 D) and with a low polar
(dimethyl carbonate, D ) 0.9 D) solvent. RT-FTIR
experiments on the diluted samples seek to determine
whether the µcalc-Rp relationship can be exploited using
polar solvents. In Figure 6 the results of these experi-
ments using laminates, to avoid evaporation of the
solvents, are shown. Unfortunately, the use of a lami-
nated experimental setup results in different light
intensity and sample thickness compared to the previ-
ous experiments with the transmission-reflection setup.
These changes result in different Rp values and as a
consequence a different correlation. However, the basic
µcalc-Rp relation remains intact. The Rp values of neat
tetrahydrofurfuryl acrylate 36 are indicative for the
changes in experimental setup as a Rp of 5.3 mol/(L s)
is found in the transmission reflection setup whereas a
Rp of 3 mol/(L s) in the transmission setup using
laminates. It should be noted that the correlation found
in the dilution experiments with inert solvents is only
valid as long as the bulk of the solution is the acrylate
monomer (up to 30% inert solvents). Still this feature
demonstrates that polymerization rate enhancements
via the addition of small amounts of an inert solvent
can be achieved contrary to the general belief that
dilution automatically leads to reductions in polymer-
ization rates.30,31
None of these possible explanations can be ruled out
so far. However, there is experimental evidence that
makes some explanations less plausible.
The activation energy of polymerization of some
acrylates has been determined using photo-DSC or RT-
FTIR.2a,c In the temperature range 30-90 °C the
maximum activation energy found is only 5 kJ /mol.
Such a low value suggests that the reaction could be
diffusion-controlled. Consequently, changes in activation
energy would only have a minor effect on the rate-
limiting process. Therefore, it seems unlikely that
further reduction of such a low value would be the cause
of the rate increase observed above. However, it should
be noted that according to pulsed laser experiments at
low temperatures (range -60 to 20 °C) the energy of
activation is higher than the above-mentioned 5 kJ /mol,
namely around 17 kJ /mol for the acrylates studied by
PLP. Still, even in this case the activation energy is
relatively low.35
As mentioned above in the introduction to this section,
the hydrogen abstraction theory offers no explanation
why the oxazolidone functional acrylate 56 is less
reactive compared to the glycerol carbonate functional
acrylate 57. The formation of a stronger solvent cage
around the propagating radical can explain the experi-
mental evidence so far, especially the rate polymeriza-
tion increase using small amounts of inert solvent.
Furthermore, this theory explains why the solvent effect
as shown in Figure 7 is only valid for low amounts of
inert solvent. The formation of a stronger solvent cage
may reduce the probability of termination (lower kt) but
normally could reduce the probability of propagation as
In summary, we have demonstrated that a direct
correlation exists between the Boltzmann-averaged

3870 J ansen et al.
Macromolecules, Vol. 36, No. 11, 2003
Ta ble 4. Kin etic P a r a m eter s a t Rp of Selected Mon om er s
well (lower kp). As the solvent cage is mainly based on
the acrylate monomer, this negative effect on kp will be
negligible in these dilution experiments using small
amounts of inert solvents, whereas the positive effect
on kt is still present. The reduction in kp at higher
dilutions, as well as the concentration effect, can
overcome the positive effect of the reduction in kt, which
eventually could lead to lower polymerization rates
employing higher dipolar solvents, obscuring the dipole
effects in diluted polymerizations.36 However, it should
be noted that, on the basis of pulsed laser experiments,
solvent effects on kp are generally small.37
A solvent cage around the propagating radical has
been suggested in order to explain the experimental
results in vinyl polymerizations.38 With aromatic sol-
vents the formation of complexes between the propagat-
ing radical and the solvent has been proposed.39 On the
basis of such a complex a kinetic model, the so-called
“reactant-solvent complex model (RSC)” has been
postulated.40
Molecular modeling calculations employing ESP charge
calculations revealed that in a high-polarity medium the
propagating radical (simulated by the methyl propionate
radical in a field with a fixed dielectric constant) is
negatively charged to a greater extent relative to the
charge in a low-polarity medium. This is demonstrated
with the following series: dimethyl carbonate D ) 0.9
D, charge -0.2049; ethylene carbonate D ) 5 D, charge
-0.2207. Although these calculations have not been
performed at the most refined level, they still give some
credibility to this theory.
compound
µcalc
kp
kt
36
HEA
54
2.05
2.93
4.54
8 × 103
1.5 × 104
7 × 103
1.2 × 106
7 × 105
7 × 104
the data presented by Decker, who found similar kp’s
but significantly different kt’s as well.
Com bin in g P r eor gan ization via Hydr ogen Bon d-
in g a n d th e Dip ole Th eor y. A significant consequence
of the fundamental explanation behind the dipole theory
implies that preorganization via hydrogen bonding and
the dipole theory can be combined and are complemen-
tary. In fact, it means that monomers capable of
hydrogen bonding should possess a higher polymeriza-
tion rate compared to monomers with the same dipole
moment but without hydrogen-bonding capability. More-
over, if the hydrogen-bonding effect can fully be at-
tributed to a change in kp, then a similar increase in Rp
as a function of polarity should be found for hydrogen-
bonding monomers as well as non-hydrogen-bonding
monomers; i.e., in Figure 5 they should be on a parallel
line, with for the same dipole moment a higher rate.
The Boltzmann-averaged dipole moments for several
acrylate monomers capable of hydrogen bonding were
calculated [ethyl-substituted amide rate 1 23.5 mol/(L
s), 4.31 D; urethane 5 rate 16.1 mol/(L s); 3.24 D;
“inverted” urethane 9 rate 15.9 mol/(L s), 3.23 D; and
urea rate 25.2 mol/(L s), 4.64 D functional ethyl acry-
lates 13 and hydroxyethyl acrylate HEA, rate 14 mol/
(L s), 2.93 D].
It is our belief that a combination of the latter two
explanations is what really occurs, as depicted in Figure
7. A stronger solvent cage and an enhanced radical
charge may be formed in a highly polar medium, and
this may account for enhanced reactivity. Further
experiments to determine the effect of solvent polarity
on the polymerization initiation (ki), propagation (kp),
and termination (kt) rate constants are required.
It should be noted that such a solvent cage or radical-
solvent complex is required for the elucidation of the
solvent effects on radical hydrogen abstraction reac-
tions. On the basis of the solvent cage theory, a good
explanation for the fact that the rates of hydrogen
abstraction are independent of the nature of the ab-
stracting radical is given.41 A prediction of the solvent
effect on the rate of hydrogen abstraction based on
Abrahams parameters, which describe the hydrogen-
donating and hydrogen-accepting abilities is possible.42
Furthermore, solvent cage effects are frequently men-
tioned as explanation of solvent effects on radical
copolymerizations.43
This solvent cage enhanced charging explanation of
the µcalc-Rp correlation has some consequences, which
we can explore further. First the kp’s of a high and low
polarity monomer should be similar whereas a signifi-
cant difference in kt should exist. Second and more
importantly, the hydrogen-bonding mechanism and the
dipole theory should be complementary.
Employing the methodology as described by Decker,11
the kinetic parameters of tetrahydrofurfuryl acrylate 36
and lactone acrylate 54 were determined (see Table 4).
The values for the kp’s and kt’s at Rp were in line with
our expectations. Similar propagation rate constants (kp)
were found. The termination rate constants (kt), how-
ever, showed a significant difference of almost 2 orders
of magnitude. These observations are comparable with
At first glance, this complementarity seems to hold.
However, the range of calculated dipoles moments of
these monomers is rather limited and varies only from
3 to 4.5 D. A suitable monomer capable of hydrogen
bonding and with a very high dipole moment is lacking.
On the basis of our knowledge gained previously, we
expected that the five-ring cyclic carbonate urethane
ethyl acrylate (synthesis is shown in Scheme 3) would
possess a very high Boltzmann-averaged dipole moment.
A µcalc of 7.5 D was obtained. Because of this very high
value, we reconsidered our method of calculating dipole
moments, especially since lithium chloride has a value
around 9 D.
To validate our method of calculation, employing the
semiempirical AM1 method, we compared the calcu-
lated Boltzmann-averaged dipole moments of ethylene
carbonate, γ-butyrolactone, dimethyl carbonate, ethyl
propionate, methacrylic acid, methyl methacrylate, n-
propyl methacrylate, and n-butyl methacrylate with
experimental values found in the literature. Moreover,
we also compared them with dipole moments calculated
using other advanced calculation methods. The first-
principle methods, which were employed, involved
Hartree-Fock (HF), density functional (BLYP and
BPW91), HF-DFT hybrid methods (B3LYP and
B3PW91), and second-order Møller-Plesset perturba-
tion theory (MP2). For the optimizations at the HF,
MP2, BLYP, BPW91, B3PW91, and B3LYP-level the
6-31G* basis sets were employed. In addition, using the
BLYP and MP2 method, the effect of the basis set size
was investigated. For γ-butyrolactone and dimethyl
carbonate geometry optimizations were also performed
using the larger 6-31G** and 6-311G** basis sets, and
single-point calculations on the 6-31G* geometries were
carried out using the 6-31G**, 6-311G**, 6-311++G**,
and 6-311+G(3df,2p) basis sets. Although the validation

Macromolecules, Vol. 36, No. 11, 2003
Sch em e 3. Syn th esis of Glycid yl Ca r bon a te Ur eth a n e Acr yla te 65a
Fast Monomers 3871
a
Reagents: (a) ethylene carbonate, NaOH; (b) COCl2, Et3N followed by Pd/C; (c) phenylchloroformate, Et3N, 0 °C; (d)
ethanolamine, -5 °C; (e) acryloyl chloride, pyridine, 0 °C.
of our calculational method was slightly hampered by
the scatter in the literature data ((0.2 D), it showed
that our method of calculating the Boltzmann-averaged
dipole moment employing AM1 performs well.
After having confirmed the method of calculating, we
embarked on the synthesis of the five-ring cyclic carbon-
ate urethane ethyl acrylate as this possesses µcalc of 7.5
D. Although 2-isocyanatoethyl acrylate might seem the
ideal precursor for this molecule, we were unable to
prepare this acrylate. Therefore, as alternative to using
2-isocyanatoethyl acrylate, we employed the synthetic
procedure shown in Scheme 3.
Glycerol carbonate was prepared in two ways. The
standard procedure starting from glycerol 60 and react-
ing with ethylene carbonate yields glycerol carbonate
62. However, this product still contains some glycerol
and higher oligomers, which are very difficult to remove
by distillation. Therefore, to obtain glycerol carbonate
62 without traces of glycerol. the synthesis started from
monoallyl-protected glycerol 61. Phosgene as well as the
less toxic di- and triphosgene can be used for the
carbonate formation, and after removal of the allyl
protecting group pure glycerol carbonate can be obtained
in 56% yield over two reaction steps. Next, the glycerol
carbonate was reacted with phenyl chloroformate, yield-
ing the active ester 63. The choice of active ester is
crucial for the success of the next reaction step. The
acidity of the liberated phenol should be as low as
possible. In this way the subsequent reaction can be
performed without the addition of base, which besides
neutralizing the liberated acid also enhanced the nu-
cleophilicity of the amine, thereby leading to side
reactions. In those cases where a tertiary amine like
triethylamine or pyridine was added, we noticed some
reaction on the cyclic carbonate yielding unwanted side
products. It should be noted that the hydroxy functional
urethane carbonate 64 is very polar and dissolves
readily in water. On the basis of alcohol 64 the synthesis
of glycidyl carbonate urethane acrylate 65 is straight-
forward, and the product can be obtained pure (starting
from 61) in a calculated 6% overall yield over five
reaction steps. In case lower purities can be tolerated,
obviously much higher yields can be obtained as distill-
ing 65 results in large product losses. Alongside this
synthesis, we used the commercially available 2-isocy-
anatoethyl methacrylate and glycidyl carbonate for the
preparation of the methacrylate analogue to 65.
F igu r e 8. Correlation between the µcalc and Rp as determined
with RT-FTIR. Closed circles, neat monomers; open circles,
mixtures of monomers; open triangles, monomers capable of
hydrogen bonding.
level of hydrogen bonding is reduced, a higher rate is
expected at room temperature. Acknowledging the µcalc
-
Rp relationship, a Rp of 44 mol/(L s) was predicted. In
line with this prediction, a Rp of 43.5 mol/L was based
on extrapolation from the methacrylate analogue. The
combined µcalc-Rp plot using all the evaluated mono-
mers is shown in Figure 8.
The scatter in Rp at low µcalc values (<3.5 D) might
be due to other effects. These effects could be for
instance polymer Tg, molecular packing phenomena due
to van der Waals interactions, or even hydrogen ab-
straction reactions, retarding the polymerization. How-
ever, it is evident that for values above 3.5 D the µcalc
-
Rp relationship is valid. Moreover, monomers capable
of both hydrogen bonding and having a high µcalc appear
to obey a similar relationship at the expected higher
polymerization rates.
The simple fact that the correlation coefficients are
the same, as illustrated with parallel lines in Figure 8
is a strong indication that both effects operate via
different mechanisms. The proposed effect on kp of
hydrogen bonding and the proposed effect on kt due to
the dipole moment are in line with this reasoning.
However, so far all the experimental evidence are only
strong indications. The final proof is measuring the
actual kinetic constants. In Table 4 the kinetic data from
two monomers not capable of hydrogen bonding, i.e., 36
and 54 and a monomer capable of hydrogen bonding
HEA at Rp are given. The trends in kt are clearly in line
with the proposed effect of the dipole moment on the
termination rate. At higher dipole moments lower
values for kt are found. The fact that hydrogen bonding
affects kp is nicely illustrated by the double value of kp
of HEA compared to 36 and 54. So the kinetic analysis
Unfortunately, the five-ring cyclic carbonate urethane
ethyl acrylate 65 was a solid at room temperature.
Consequently, the maximum rate (38 mol/(L s)) was
determined at 50 °C. Since at higher temperatures the

3872 J ansen et al.
Macromolecules, Vol. 36, No. 11, 2003
1991, 27, 881. (f) Moussa, K.; Decker, C. J . Polym. Sci., Part
A: Polym. Chem. 1993, 31, 2197. (g) Decker, C.; Elzaouk, B.
Eur. Polym. J . 1995, 31, 1155. (h) Decker, C.; Elzaouk, B.;
Decker, D. J . Macromol. Sci., Pure Appl. Chem. 1996, A33,
173.
corroborate the theories presented here: hydrogen
bonding causes rate enhancement by increasing kp, and
a higher dipole moment causes a rate enhancements by
decreasing kt.
(12) Andrzejewska, E.; Andrzejewski, M. J . Polym. Sci., Part A:
Polym. Chem. 1998, 36, 665.
Con clu sion
(13) (a) Guymon, C. A.; Hoggan, E. N.; Clark, N. A.; Rieker, T.
P.; Walba, D. M.; Bowman, C. N. Science 1997, 275, 57. (b)
Guymon, C. A.; Bowman, C. N. Macromolecules 1997, 30,
5271. (c) Lester, C. L.; Colson, C. D.; Guymon, C. A.
Macromolecules 2001, 34, 4430.
(14) (a) J ansen, J . F. G. A.; Dias, A. A.; Dorschu, M.; Coussens,
B. Polym. Prepr. 2001, 42, 769. (b) J ansen, J . F. G. A.; Dias,
A. A.; Dorschu, M.; Coussens, B. Macromolecules 2002, 35,
7529. (c) J ansen, J . F. G. A.; Dias, A. A.; Dorschu, M.;
Coussens, B. Patent WO 02/42383.
In conclusion, we have given evidence that Rp is
dominated by two molecular factors: hydrogen bonding
and the Boltzmann-averaged dipole moment, which are
complementary to each other. Hydrogen bonding affects
Rp via preorganization of the monomers whereas the
Boltzmann-averaged dipole moment (µcalc) may affect Rp
due to a stronger solvent cage combined with increased
charge at the propagating radical. This understanding
of acrylate reactivity at the molecular level enables the
rational design of monomers with a defined reactivity
as opposed to the trial-and-error strategy applied so far.
Moreover, the effect of the addition of small amounts
of polar solvents on Rp is demonstrated and could prove
useful in practical systems. In fact, solvents can be used
to enhance the reactivity of the acrylate monomers.
(15) Dias, A. A.; Hartwig, H.; J ansen, J . F. G. A. Surf. Coat. Int.,
J OCCA 2000, 83, 382.
(16) (a) Decker, C.; Moussa, K. Polym. Prepr. 1988, 29, 516. (b)
Decker, C.; Moussa, K. Makromol. Chem. 1988, 189, 2381.
(c) Decker, C.; Moussa, K. Eur. Polym. J . 1990, 26, 393. (d)
Decker, C.; Moussa, K. Makromol. Chem. 1990, 191, 963. (e)
Decker, C. Macromolecules 1990, 23, 5217. (f) Decker, C.;
Moussa, K. J . Coat. Technol. 1990, 62, 55. (g) Decker, C. J .
Polym. Sci., Part A: Polym. Chem. 1992, 30, 913. (h) Decker,
C. Spec. Publ.sR. Soc. Chem. 1993, 125, 32 and refs 11g and
11h.
(17) The formation of gel particles in butyl acrylate has been
described in: Plessis, C.; Arzamendi, G.; Leiza, J . R.; Schoon-
brood, H. A. S.; Charmont, D.; Ausa, J . M. Macromolecules
2000, 33, 4.
Ack n ow led gm en t. The management of DSM Re-
search, DSM Coating Resins, and DSM Desotech are
kindly acknowledged for their permission to publish this
work.
(18) (a) Computational results obtained using software programs
from Molecular Simulations force field calculations were done
with the Discover program, using the CVFF force-field,
semiempirical calculations were done with the MOPAC 6.0
program. (b) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.;
Scuseria, G. E.; Robb, M. A.; Cheeseman, J . R.; Zakrzewski,
V. G.; Montgomery, J . A., J r.; Stratmann, R. E.; Burant, J .
C.; Drapprich, S.; Millam, J . M; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J .; Barone, V.; Cossi, M.;
Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford,
S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov,
B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J .; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challa-
combe, M.; Gill, P. M. W.; J ohnson, B.; Chen, W.; Wong, M.
W.; Andres, J . L.; Gonzalez, C.; Head-Gordon, M.; Replogle,
E. S.; Pople, J . A. Gaussian 98, revision A.5; Gaussian, Inc.:
Pittsburgh, PA, 1998. See also: Scott, A. P.; Radom, L. J .
Phys. Chem. 1996, 100, 16502.
(19) The naming of the compounds is according to the following
scheme: alkyl group-second functionality-alkylbridge-
acrylate. In case of urethanes also the atoms, i.e., N and O,
which connects to the alkyl groups are mentioned. Using this
naming methodology, the changes in the second functionality
can be directly compared with their effect on the rate of
polymerization.
(20) (a) Krimm, H.; Buysch, H.-J . European Patent EP 57360,
1998. (b) Buysch, H.-J .; Fengler, G.; Neumann, K.-H.;
Wagner, P. German Patent DE 19625265, 1996. (c) Buysch,
H.-J .; Fengler, G.; Neumann, K.-H.; Wagner, P.; Melchiors,
M.; Hovestadt, W. European Patent EP 0816354, 1996.
(21) (a) Pelet, S.; Yoo, J .-W.; Moulongui, Z. J . High-Resolution
Chromatogr. 1999 22, 276. (b) Moulongui, Z.; Yoo, J .-W.;
Gahen, C.-A.; Gaset, A.; Vermeersch, G. European Patent
EP739888, 1996. (c) Kawabata, O.; Tanimoto, F.; Inoe, Y.
J apanese Patent J P0600910, 1993. (d) Teles, J .-H.; Rieber,
N.; Harder, W. European Patent EP 582201, 1993. (e)
Rokichi, G.; Kuran, W.; Pogorzelscha-Marciniak, B. Monatsch.
Chem. 1984, 115, 205. (f) Rokichi, G.; Kuran, W. Bull. Chem.
Soc. J pn. 1984, 57, 1662.
Refer en ces a n d Notes
(1) (a) Nie, J .; Rabek, J . F.; Linden, L.-A. Polym. Int. 1999, 48,
129. (b) Narayanan, V.; Scranton, A. B. Trends Polym. Sci.
1997, 5, 415. (c) Lovell, L. G.; Stansburry, J . W.; Syrpes, D.
C.; Bowman, C. N. Macromolecules 1999, 32, 3913 and
references therein. (d) Kloosterboer, J . G. Adv. Polym. Sci.
1988, 84, 1.
(2) (a) Tryson, G. R.; Schultz, A. R. J . Polym. Sci., Polym. Phys.
Ed. 1979, 17, 2059. (b) Decker, C.; Elzaouk, B.; Decker, D. J .
Macromol. Sci., Pure Appl. Chem. 1996, A33, 173 and
references therein. (c) Doornkamp, A. T.; Tan, Y. Y. Polym.
Commun. 1990, 31, 362. (d) Lecamp, L.; Youssef, B.; Bunel,
C.; Lebaudy, P. Polymer 1999, 40, 1403. (e) Khudyakov, I.
V.; Legg, J . C.; Purvis, M. M.; Overton, B. J . Ind. Eng. Chem.
Res. 1999, 38, 3353. (f) Selli, E.; Bellobono, I. R. In Radiation
Curing in Polymer Science and Technology; Fouassier, J . P.,
Rabek, J . F., Eds.; Elsevier: London, Vol. 3, p 1. (g) Decker,
C.; Masson, F.; Bianchi, C. Polym. Prepr. 2001, 42, 304. (h)
Yamada, B.; Azukizawa, M.; Yamazoe, H.; Hill, D. J . T.;
Pomery, P. J . Polymer 2000, 41, 5611. (i) Seno, M.; Fukui,
T.; Hirano, T.; Sato, T. J . Polym. Sci., Part A: Polym. Chem.
2000, 38, 4264. (j) Doetschman, D. C.; Mehlenbach, R. C.;
Cywar, D. Macromolecules 1996, 29, 1807.
(3) (a) Anseth, K. S.; Bowman, C. N. Polym. React. Eng. 1993,
1, 499. (b) Anseth, K. S.; Wang, C. M.; Bowman, C. N.
Macromolecules 1994, 27, 650. (c) Young, J . S.; Bowman, C.
N. Macromolecules 1999, 32, 6073.
(4) Wen, M.; McCormick, A. V. Macromolecules 2000, 33, 9247.
(5) Goodner, M. D.; Bowman, C. N. Macromolecules 1999, 32,
6552.
(6) Berchtold, K. A.; Hacioglu, B.; Lovell, L. G.; Nie, J .; Bowman,
C. N. Macromolecules 2001, 34, 5103.
(7) (a) Hutchinson, J . B.; Anseth, K. S. Polym. Prepr. 2000, 41,
1326. (b) Lu, H.; Lovell, L. G.; Bowman, C. N. Macromolecules
2001, 34, 8021.
(8) (a) Elliot, J . E.; Bowman, C. N. Macromolecules 1999, 32,
8621. (b) Elliot, B. J .; Willis, W. B.; Bowman, C. N. Macro-
molecules 1999, 32, 3201. (c) Elliot, B. J .; Scranton, A. B.;
Cameron, J . H.; Bowman, C. N. Chem. Mater. 2000, 12, 633.
(9) (a) Elliot, J . E.; Bowman, C. N. Macromolecules 2001, 34,
4642. (b) Lovell, L. G.; Stansburry, J . W.; Syrpes, D. C.;
Bowman, C. N. Macromolecules 1999, 32, 3913.
(22) Buermann, S.; Paquet, D. A., J r.; McMinn, J . H.; Hutchinson,
R. A. Macromolecules 1997, 30, 194.
(23) Scherzer, T.; Decker, U. Polymer 2000, 41, 7681.
(24) Pham, Q.-T.; Petiaud, R.; Waton, H.; Llauro, M.-F. Proton
and Carbon NMR Spectra of Polymers; J . Wiley & Sons: New
York, 1984; Vol 3 and references cited therein.
(25) Berchtold, K. A.; Nie, J .; Elliot, J . E.; Hacioglu, B.; Luo, N.;
Trotter, A. J . N.; Stansburry, J . W.; Bowman, C. N. Proc.
Radtech Europe 2001, 265.
(10) Berchtold, K. A.; Lovell, L. G.; Nie, J .; Hacioglu, B.; Bowman,
C. N. Polymer 2001, 42, 492 and references therein.
(11) (a) Decker, C.; Moussa, K. Polym. Mater. Sci. Eng. 1989, 60,
547. (b) Decker, C.; Moussa, K. Makromol. Chem. Rapid
Commun. 1990, 11, 159. (c) Decker, C.; Moussa, K. Makromol.
Chem. 1990, 192, 507. (d) Decker, C.; Moussa, K. Eur. Polym.
J . 1991, 27, 403. (e) Decker, C.; Moussa, K. Eur. Polym. J .

Macromolecules, Vol. 36, No. 11, 2003
Fast Monomers 3873
(26) Employing organogels the Rp of lauryl acrylate was only
increased by 15%, although a solid gel was formed. For
organogelators see: (a) de Loos, M.; van Esch, J .; Stokroos,
I.; Kellogg, R. M.; Feringa, B. L. J . Am. Chem. Soc. 1997,
119, 12675. (b) van Esch, J .; DeFeyter, S.; Kellogg, R. M.;
DeSchrijver, F.; Feringa, B. L. Chem.sEur. J . 1997, 3, 1238.
(c) de Loos, M.; van Esch, J .; Kellogg, R. M.; Feringa, B. L.
Angew. Chem., Int. Ed. 2001, 40, 613.
(35) Using pulsed laser polymerization higher Ea are found. See:
van Herk, A. M. J . Macromol. Sci., Rev. Macromol. Chem.
Phys. 1997, C37, 633.
(36) Olaj, F. J .; Schnoll-Bitai, I. Monatsch. Chem. 1999, 130, 731.
(37) (a) Davis, T. P.; O’Driscoll, K. F.; Piton, M. C.; Winnink, M.
A. Macromolecules 1989, 22, 2785. (b) Beuermann, S.; Bubak,
M.; Russel, G. T. Macromol., Rapid Commun. 1994, 15, 351.
(c) Morrison, B. R.; Piton, M. C.; Winnink, M. A.; Gilbert, R.
G.; Napper, D. H. Macromolecules 1993, 26, 4368. (d) Coote,
M.; Davis, T. P.; Klumperman, B.; Monteiro, M. J . Macromol.
Sci., Rev. Macromol. Chem. Phys. 1998, C38, 567. (e) Also in
ATRP the kp is solvent independent. See: Chambard, G.;
Klumperman, B.; German, A. L. Macromolecules 2002, 35,
3420.
(27) For an excellent review and discussion of the gel effect see:
O’Niel, G. A.; Wisnudel, M. B.; Torkelson, J . M. Macro-
molecules 1996, 29, 7477.
(28) Tedder, J . M.; Walton, J . C. Tetrahedron 1980, 36, 701.
(29) Tedder, J . M. Angew. Chem., Int. Ed. Engl. 1982, 21, 401.
(30) It should be noted that solvent effects in free radical
polymerisations are hardly understood. See: Comprehensive
Polymer Science; Allen, G., Bevington, J . C., Eds.; Perga-
mon: Oxford, 1989; Vol. 3. For examples of solvent effects in
homopolymerizations see: (a) Matyjaszewski, K.; Nakagawa,
Y.; J asciezek, C. B. Macromolecules 1998, 31, 1535. (b)
Couvreur, L.; Piteau, G.; Castignolles, P.; Tonge, M.; Coutin,
B.; Charleux, B.; Vairon, J .-P. Macromol. Symp. 2001, 174,
197. (c) Alvarez, J .; Encinas, M. V.; Lissi, E. Macromol. Chem.
Phys. 1999, 200, 2411. (d) Wang, B.; Xie, S.; Cao, M.; Chin.
J . Polym. Sci. 1987, 5, 141.
(31) In atom transfer polymerizations a similar effect, i.e., higher
rates at a higher sovent polarity has been observed see: (a)
Chambard, G.; Klumperman, B.; German, A. L. Macromol-
ecules 2000, 33, 4417. (b) Kumar Nanda, A.; Matyjaszewski,
K. Macromolecules 2003, 36, 599.
(32) For solvents effects on initiation see: Turro, N. J .; Kleinman,
M. H.; Karatekin, E. Angew. Chem., Int. Ed. 2000, 39, 4436.
(33) The effect of a stronger solvent cage on the rate of termination
has already been postulated: Mayer, G.; Schultz, G. V.
Makromol. Chem. 1973, 173, 101.
(38) McKenna, T. F.; Villanueva, A.; Santos, A. M. J . Polym. Sci.,
Part A: Polym. Chem. 1999, 37, 571.
(39) (a) Kamachi, M.; Liaw, D.; Nozakura, S. Polym. J . 1979, 11,
921. (b) Kamachi, M. Adv. Polym. Sci. 1981, 38, 55. (c)
Kamachi, M. Adv. Polym. Sci. 1987, 82, 209.
(40) (a) Czerwinski, W. K. Makromol. Chem. Theory Simul. 1993,
2, 577. (b) Czerwinski, W. K. Makromol. Chem. 1991, 192,
577. (c) Czerwinski, W. K. Makromol. Chem. 1992, 193, 359.
(d) Czerwinski, W. K. Makromol. Chem. 1993, 194, 3015.
(e) Czerwinski, W. K. Macromolecules 1995, 28, 5405. (f)
Czerwinski, W. K. Macromolecules 1995, 28, 5411. A serious
disadvantage of this model is the large number of optimizable
parameters.
(41) Lucarini, M.; Pedulli, G. F.; Valgimigli, L. J . Org. Chem. 1998,
63, 4497.
(42) Snelgrove, D. W.; Lusztyk, J .; Banks, J . T.; Mulder, P.; Ingold,
K. U. J . Am. Chem. Soc. 2001, 123, 469.
(43) (a) Matsumoto, A.; Mohri, Y. J . Polym. Sci., Part A: Polym.
Chem. 1999, 37, 2803. (b) Plochocvka, K. J . Macromol. Sci.,
Rev. Macromol. Chem. 1981, C20, 67.
(34) In radical reactions the ionic nature of the attacking radical
is important, See: Citterio, A.; Arnoldi, A.; Minisci, F. J . Org.
Chem. 1979, 44, 2674.
MA021785R