Investigations of the reactivity of "kick-started" oxetanes in photoinitiated cationic polymerization

Article abstract of DOI:10.1002/pola.27479

The ability of certain alkyl substituted epoxides to accelerate the photoinitiated cationic ring-opening polymerizations of oxetane monomers by substantially reducing or eliminating the induction period altogether has been termed by us "kick-starting." In

Full text of DOI:10.1002/pola.27479

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Investigations of the Reactivity of “Kick-Started” Oxetanes in
Photoinitiated Cationic Polymerization
James V. Crivello
Department of Chemistry and Chemical Biology, Rensselaer Polytechnic Institute, Troy, New York, 12180
Correspondence to: J. V. Crivello (E-mail: crivej@rpi.edu)
Received 19 October 2014; accepted 16 November 2014; published online 00 Month 2014
DOI: 10.1002/pola.27479
ABSTRACT: The ability of certain alkyl substituted epoxides to
accelerate the photoinitiated cationic ring-opening polymeriza-
tions of oxetane monomers by substantially reducing or elimi-
nating the induction period altogether has been termed by us
“kick-starting.” In this communication, the rates of photopoly-
merization of several model “kick-started” oxetane systems
were quantified and compared with the analogous biscycloali-
phatic epoxide monomer, 3,4-epoxycyclohexylmethyl 30,40-
epoxycyclohexanecarboxylate (ERL). It has been found that the
“kick-started” systems undergo photopolymerization at rates
that are at least two-fold faster than ERL. These results suggest
that “kick-started” oxetanes could replace ERL in many applica-
tions in which high speed ultraviolet induced crosslinking pho-
C
V
topolymerizations are carried out.
2014 Wiley Periodicals,
Inc. J. Polym. Sci., Part A: Polym. Chem. 2014, 00, 000–000
KEYWORDS: crosslinking; cationic polymerization; kinetics; pho-
topolymerization; epoxides
polymerizations and this has been an impediment to the
growth and use of these monomers. The fundamental reasons
that underlie the sluggish reactivity of oxetane monomers in
photoinitiated cationic polymerization have remained elusive
and have not been fully explored until recently.
INTRODUCTION Work in this laboratory over the past two
decades has been focused on the photoinitiated cationic
ring-opening polymerization of epoxide monomers. These
photopolymerizations have found a wide range of commer-
cial uses in thin film applications such as coatings for metals,
plastics, glass and wood, pressure sensitive adhesives, mark-
ing inks for electronic components, and in additive manufac-
turing [three-dimensional (3D) imaging]. However, epoxide
photopolymerizations have a number of drawbacks. The
most useful epoxide monomers are those that are prepared
by the epoxidation of carbon–carbon double bond containing
substrates with peroxides. Due to the well-known1 hazards
associated with peroxides, the commercial availability of
epoxides by this method is currently restricted to only a few
monomers that command rather high prices. Further, many
epoxides also possess considerable toxicological issues, espe-
cially with respect to undesirable eye and skin irritation.
The photoinitiated acid-catalyzed cationic ring-opening poly-
merization of oxetane monomers proceeds by the same
mechanism as epoxides5 and is depicted in Scheme 1. For
purposes of clarity, this mechanism is depicted using the
unsubstituted compound oxetane (trimethylene oxide). The
photolysis of a photoacid generator such as a diaryliodonium
or triarylsufonium salt produces a strong Brønsted acid
(eq 1) that protonates the oxygen atom of the oxetane ring
(eq 2). Then, ring-opening (eq 3) occurs by a SN2 attack of
the oxygen atom of a neutral monomer on one of the carbon
atoms adjacent to the protonated oxygen atom of the mono-
mer (secondary oxonium ion), 1, to give tertiary oxonium
ion, 2. This latter reaction is fast in the case of an epoxide,
but considerably slower for an oxetane. Previously, Sasaki
et al.6,7 have pointed out that the oxygen ring atom of an
oxetane is considerably more basic than the oxygen atom in
an epoxide. This leads to a higher energy barrier for the
ring-opening of the protonated oxetane. Subsequent SN2
attack by monomer molecules on the respective tertiary oxo-
nium ion, 2, (eq 4) is rapid due to relief of the ring-strain in
the oxonium ion. The third step of the mechanism shown in
Scheme 1 is the rate-determining step. Since each of the
Oxetanes are closely related compounds to epoxides and
share many of the same desirable characteristics. Moreover,
the synthetic methods used for the preparation of oxetane
monomers do not involve epoxidation. Instead, the prepara-
tive methods rely on simple, direct base-mediated ring-clo-
sure2 or thermally induced carbon dioxide extrusion
reactions.3 Oxetanes are generally considerably less toxic and,
consequently, more easily handled.4 Despite their similar ring-
strain, oxetanes have been reported by formulators to be con-
siderably less reactive than epoxides in photoinitiated cationic
C
V
2014 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
1

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
EXPERIMENTAL
Materials
1,2,8,9-Limonene dioxide [limonene dioxide (LDO), a mix-
ture of isomers] was provided as a sample by Arkema, Inc.,
Grand Rapids, MI, and purified by fractional vacuum distil-
lation before use. 2,2,3,-Trimethyloxirane (2,3-epoxy-2-
methybutane) was purchased from Acros Organics, Wal-
tham, MA. 3,4-Epoxycyclohexylmethyl 30,40-cyclohexylcar-
boxylate (ERL-4221E, abbreviated ERL) was purchased
from the Union Carbide Corporation (now Dow Chemical
Corp., Midland, MI). 3-Ethyl-3(2-ethylhexyloxymethyloxe-
tane [EHOXT]), 3-ethyl-3(2-phenoxymethyloxetane [POX]),
and bis{[(1-ethyl(3-oxetanyl)]methyl} ether (DOX), were
gratefully received as gifts from the Toagosei Chemical
Company, Nagoya, Japan. POX and DOX were purified prior
to use by fractional vacuum distillation (POX bp 117–118/
0.6 mmHg). A sample of Irgacure 651 was kindly supplied
by the Ciba Specialty Products Corp., Basel, Switzerland.
1,2-Cyclohexene oxide was purified by distillation from cal-
cium hydride. Other reagents used in this work were
obtained from the Aldrich Chemical Company, Milwaukee,
WI, and used as received without further purification. The
diaryliodonium salt8 and triarylsulfonium9 salt and dialkyl-
phenacylsulfonium salt10 photoinitiators were prepared as
described previously. In this article, three of these photoini-
tiators were employed and we have developed shorthand
designations for these compounds. For example, IOC-8 SbF6
refers to (4-n-octyloxyphenyl)phenyliodonium hexafluoroan-
timonate, having the structure shown below in which an n-
octyloxy group is attached to one of the phenyl groups in
the 4-position, while SbF6 denotes the hexafluoroantimo-
nate anion. In a similar manner, SOC-10 SbF6 refers to S(4-
n-decyloxyphenyl)-S,S-diphenylsulfonium hexafluoroantimo-
nate and SOC-10 PF6 to S(4-n-decyloxyphenyl)-S,S-diphenyl-
sulfonium hexafluorophosphate with the structures
indicated below. DAPSC1,C8 SbF6 is a shorthand notation
for S-n-octyl-S-methyl-S-phenacylsulfonium hexafluoroan-
timonate.
SCHEME 1 Mechanism of the photoinitiated cationic polymer-
ization of oxetanes.
propagation steps involves SN2 reactions, substitution at the
2 and 4 positions of the oxetane ring would be expected to
have a large impact on slowing the rates of those reactions.
For that reason, the work described in this communication
has been specifically restricted to 3-mono- and 3,3-disubsti-
tuted oxetanes.
There are several consequences of the mechanistic scenario
presented in Scheme 1 and they are further exemplified in
the optical pyrometry (OP) plots shown in Figure 1. In the
photoinitiated cationic polymerization of a simple, highly
strained epoxide monomer such as 1,2-cyclohexene oxide,
polymerization begins almost instantaneously at the start of
the ultraviolet (UV) irradiation as the Brønsted acid is pro-
duced by photolysis of the photoinitiator. In the case of
cyclohexene oxide, the photolysis of the photoinitiator is the
rate-determining step. For that reason, those factors that
directly affect the photolysis of the photoinitiator including
the quantum yield and UV irradiation intensity have a large
impact on the rate of photopolymerization. The shape of the
OP curve for cyclohexene oxide given in Figure 1 is compli-
cated in the latter stages by the vaporization of the mono-
mer due to the highly exothermic polymerization. Under the
same conditions, the oxetane monomer, 3-ethyl-3(2-ethylhex-
yloxymethyl)oxetane displays quite different behavior. As
irradiation proceeds, the protonated secondary oxonium ion
that is formed does not react further with monomer to yield
polymer. Instead, the concentration of the protonated oxe-
tane species builds up in the reaction mixture and this is
manifested by an induction period. Once the onset of the
oxetane polymerization takes place, a rapid, autoaccelerated
exothermic ring-opening polymerization occurs producing a
sharp rise in sample temperature.
The presence of the oxetane induction period is highly unde-
sirable and makes these monomers unsuitable for many in-
line, high-speed coating, printing, and bonding applications
that are the major uses of current photopolymerization tech-
nology. Accordingly, it was the purpose of this investigation
to explore means of eliminating the induction period in the
photoinitiated cationic polymerizations of oxetane monomers
and oxetane-functional oligomers.
FIGURE 1 Comparisons of the photoinitiated cationic polymer-
izations of 1,2-cyclohexene oxide and 3-ethyl-3(2-ethylhexylox-
ymethyl)oxetane using 2.0 wt
% (0.3 and 0.7 mol %
respectively) of S(4-n-decyloxyphenyl)-S,S-diphenylsulfonium
hexafluoroantimonate (SOC-10 SbF6).
2
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
In a recent article from this laboratory,17 we have described
the marked acceleration of the photoinitiated cationic ring-
opening polymerization of 3-mono- and 3,3-disubstituted
oxetanes by certain alkylated epoxide compounds. It was
observed that, due to steric hindrance, 2,2-, 2,2,3-tri-, and
2,2,3,3-tetraalkyl substituted epoxides either undergo facile
rearrangements to aldehydes or ketones or very inefficient
homopolymerization in the presence of cationic initiators. In
contrast, when these same epoxides were present in small
concentrations in mixtures with oxetanes, they were shown
to markedly accelerate the cationic ring-opening polymeriza-
tions of those monomers. Investigations showed that the
acceleration of the rate of the oxetane polymerization by
these specific epoxides was due to a basic change in the
mechanism of the initiation step of the polymerization.
Depicted in Scheme 2 is a sequence of reaction steps we
have proposed to explain the experimental results. Attack
by the photogenerated Brønsted acid on 2,2,3-trimethyloxir-
ane results in ring-opening to give the tertiary carbocation, 3
OP Characterization of Oxetane and Epoxide
Photopolymerizations
We have previously described the analytical techniques and
apparatus used in this laboratory for OP.11,12 Samples for OP
kinetic analysis were prepared by sandwiching a liquid mono-
mer containing the designated photoinitiator between two
thin (12.5 mm) films of fluorinated poly(ethylene-co-
propylene) (DuPont FEP thermoplastic film) using a thin,
open polyester mesh as a spacer. The samples were mounted
in plastic 2 cm 3 2 cm slide frames and then inserted into
the sample holder for analysis. The average thickness of the
samples was 0.912 mm. Irradiation with UV light was accom-
plished using a UVEXS Model SCU-110 mercury arc lamp
(Sunnyvale, CA) equipped with a liquid optic cable. The liquid
optic cable served as a light filter passing UV light of wave-
lengths greater than 300 nm, but blocks both shorter wave-
lengths as well as wavelengths in the infrared region. UV
irradiation intensities were measured using a Control Cure
Radiometer (UV Process Supply, Chicago, IL). Several kinetic
runs were performed for each photopolymerizable system
and the results reported in this article were the average of at
least three kinetic runs. Typically, the reproducibility of the
kinetic data was 65%. All kinetic studies were conducted at
ambient laboratory temperature (25 ꢀC–28 ꢀC) unless other-
wise noted. Photoinitiator concentrations are given in either
or both mol % and wt % depending on the given experiment.
RESULTS AND DISCUSSION
A common strategy for increasing the apparent reactivity of
a monomer is to copolymerize it with a more reactive mono-
mer. Most of the efforts directed at accelerating the photopo-
lymerization of oxetane monomers have been undertaken
using 3,4-epoxycyclohexylmethyl 30,40-epoxycyclohexanecar-
boxylate (ERL).13–16 Shown in Figure 2 are plots of the
induction periods of the mono- and bisoxetane monomers;
POX (3-ethyl-3(2-phenoxymethyloxetane) and DOX (bis{[(1-
ethyl(3-oxetanyl)]methyl} ether) as a function of the molar
content of ERL present in the mixture. From these two stud-
ies, it may be concluded that the reduction in the induction
period of an oxetane monomer is directly proportional to the
amount of ERL present in the mixture. Thus, substantial
amounts of ERL must be added to appreciably shorten the
induction period. However, in no case studied, was the
induction period eliminated. Based on these results, it is
understandable why oxetanes have been employed as modi-
fiers of epoxy resins and not the other way around.
FIGURE 2 (A) Effect of ERL on the induction period of the pho-
toinitiated cationic polymerization of POX (2.0 wt %, 0.3 mol %
SOC-10 SbF6 as photoinitiator, light intensity, 2910 mJ/cm2
min). (B) Effect of ERL on the induction period of the photoiniti-
ated cationic polymerization of DOX (2.0 wt %, 0.3 mol % SOC-
10 SbF6 as photoinitiator, light intensity, 1350 mJ/cm2 min).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
3

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Shown in Figure 3 is a comparison of the photopolymeriza-
tions of 3-ethyl-3(2-ethylhexyloxy)methyloxetane carried out
in the presence and absence of LDO. The rapid photopolyme-
rization without a substantial induction period is characteris-
tic of “kick-started” oxetane photopolymerizations carried
out in the presence of a small amount of the diepoxide. This
is in stark contrast with the results shown in Figure 2(A,B)
obtained with ERL as a modifier and is indicative of a major
change in the mechanism of the polymerization.
OP12 was used to monitor the progress of the photopolyme-
rizations shown in Figures 1 and 3 in real time. Real-time
infrared spectroscopy (RTIR) was found to be insensitive for
these photopolymerizations due to the lack of strong unob-
scured absorption bands in the IR spectra of oxetane mono-
mers. The OP analytical method is useful for comparing the
reactivity of different monomer types. Given in Figure 4(A,B)
are two comparative OP studies of the respective cationic
photopolymerizations of a 4:1 molar “kick-started” mixture
of DOX and LDO with the photopolymerization of ERL. Both
polymerizations were conducted using a light intensity of
1174 mJ/cm2 min with 2.0 mol % of SOC-10 SbF6 as the
photoinitiator. A visual inspection of the OP temperature ver-
sus time profiles of the two different photopolymerizations
(solid lines) led to the conclusion that the “kick-started” sys-
tem was considerably more reactive than ERL.
SCHEME 2 Proposed mechanism of the acceleration of the
photoinitiated cationic ring-opening polymerization of oxetane
monomers by epoxides.
(eq 5). This species subsequently adds to the oxetane to pro-
duce the tertiary oxonium ion, 4, (eq 6) that initiates the fur-
ther polymerization of the oxetane monomer (eq 7).
Comparing the mechanism shown in Scheme 1 with that
depicted in Scheme 2, it can be seen that in the latter mech-
anism, the direct alkylation of the monomer by the carboca-
tion, 3, to form the highly reactive tertiary oxonium species,
4 replaces the slow step of Scheme 1 (eq 3) involving the
formation of a secondary oxonium ion by protonation of the
oxetane monomer. Since we have previously shown18 that
the acceleration and reduction of the induction period in the
Previously,18 we have demonstrated that the temperature
versus time plots obtained by OP can be converted by inte-
gration to curves that are very closely related to conversion
versus time profiles obtained by Fourier transform RTIR (FT-
RTIR). Moreover, the values of the slopes of the early por-
tions of the integrated OP curves give values which are
numerically nearly identical to Rp/[M] values for the photo-
polymerization of the same monomer obtained by FT-RTIR.
Also shown in Figure 4(A,B), respectively, are the integrated
OP curves (dashed lines) for the respective photoinitiated
“kick-started” DOX/LDO and ERL polymerizations. The rate
photopolymerization of
a “kick-started” system is not
affected by rather large variations in the oxetane–oxirane
molar ratio in the mixture, but, instead, are sensitive to var-
iations in photoinitiator concentration and light intensity, we
suggest that in such systems, the rate-determining step is
photogeneration of the Brønsted acid (eq 1). Acceleration of
the cationic photopolymerization of all 3-mono- or 3,3-disub-
stituted oxetanes investigated has been observed with a vari-
ety of 2,2-di-, 2,2,3-tri-, and 2,2,3,3-tetraalkyl-substituted
epoxides. In all cases, these epoxides are capable of under-
going proton-induced ring-opening to generate tertiary alkyl
carbocations capable of alkylating the oxetane monomer. In
the special cases of diepoxides such as LDO and terpinolene
diepoxide, both epoxide groups can function as reactive sites
for initiation of oxetane polymerization
FIGURE 3 Photoinitiated cationic polymerization 3-ethyl-3(2-
ethylhexyloxy)methyloxetane (EEHOXT) conducted alone and
in the presence of 8:1 and 4:1 molar ratios with LDO using 2.0
wt % (0.3 mol %) of S(4-n-decyloxyphenyl)-S,S-diphenylsulfo-
nium hexafluoroantimonate (SOC-10 SbF6) as a photoinitiator
at a light intensity of 1400 mJ/cm2 min.
4
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
and the fꢀormation of DOX3.21 DOX is a colorless, high boiling
(bp 119 C/5 mmHg) low viscosity (12.8 mPa s) liquid com-
pound with almost no detectable odor. DOX is also notable
for its very low toxicity characteristics.22
Replacement of LDO with 2,2,3-trimethyloxirane in a 1:4
molar mixture with DOX gave complimentary results as
shown in Figure 5. We have observed that trisubstituted,
open-chain epoxides such as 2,2,3-trimethyloxirane display
excellent ability to accelerate the overall rates of oxetane
monomers. Similarly, a number of terpene epoxides (e.g., 1,2-
limonene oxide, alpha-pinene oxide and carene oxide) that
possess 2,2,3-trialkylepoxy groups also function well as pho-
topolymerization rate accelerators.17 The compositions
reported in this manuscript have not been optimized with
respect to the amounts of oxetane and epoxide modifiers.
We have definitively shown that that a 16:1 molar ratio of
DOX to epoxide accelerator both increases the rate of oxe-
tane polymerization and also eliminates the induction period.
On a weight basis, the inclusion of as little as 3% of LDO is
sufficient to “kick-start” oxetane photopolymerizations.
FIGURE 4 OP comparison of the cationic polymerizations of A:
an 4:1 molar “kick-started” mixture of DOX and LDO with B:
ERL containing 2.0 wt % (0.3 mol %) of SOC-10 SbF6 (solid
curve, time vs. temperature plots; dashed line, time vs. conver-
sion plots).
of the photopolymerization of the former “kick-started” oxe-
tane is more than twice that of ERL. Under very similar con-
ditions, the Rp value for the photopolymerization of an 8:1
molar mixture of DOX and LDO was found to be 3.2. It
should be noted that integrated curves in both of the studies
shown are based on normalized values for the conversion
assuming complete conversion of either oxetane or epoxide
functional curves. While this assumption appears to be valid
for the oxetane system, under the conditions of these experi-
ments, the epoxide conversion for LDO falls considerably
short of 100%. Nevertheless, the Rp values taken from the
slopes of the initial portions of the integrated temperature
versus time OP curves are a reasonable measure of the reac-
tivity of that system.
As previously mentioned, the high reactivity of “kick-started”
oxetane systems was achieved by a change in mechanism
that eliminates the slow, rate-determining step in their
ring-opening polymerizations. This results in a shift of the
kinetic rate-determining step from equation 3 in Scheme 1
to equation 1 in Scheme 2, that is, the photolysis of the
onium salt photoinitiator. Consequently, factors that directly
The bisoxetane monomer, DOX, appears to be an excellent
offset for ERL. This monomer is prepared by the straightfor-
ward two-step synthesis scheme depicted in equations 8 and
9.19 Ditrimethylolpropane obtained as a byproduct of trime-
thylolpropane (TMP).20 TMP is synthesized by the base-
catalyzed sequential aldol condensation and Canazzaro reac-
tion of formaldehyde with butyraldehyde and is widely used
in the production of commodity polyesters and polyur-
ethanes. Condensation of ditrimethylolpropane with diethyl-
carbonate as shown in eq 8 gives the bis(cyclic carbonate),
5. Thermolysis of 5 (eq 9) in the presence of catalysts such
as lithium chloride results in the extrusion of carbon dioxide
FIGURE 5 Photopolymerization study of a 4:1 mixture DOX
and 2,2,3-trimethyloxetane in the presence of 2.0 wt % (0.3 mol
%) SOC-10 SbF6 at a light intensity of 1430 mJ cm2 min (solid
curve, time vs. temperature plot; dashed line, time vs. conver-
sion plot).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
5

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 8 Reduction in the induction period of an 4:1 molar
mixture of DOX and LDO by the addition of a free radical pho-
toinitiator (Irgacure 651; light intensity 2320 mJ/cm2 min; pho-
toinitiator concentrations given in wt %).
FIGURE 6 OP comparative study of the photopolymerization of
an 8:1 molar mixture of DOX and LDO with 0.3 mol % SOC-10
SbF6 and with 0.3 mol % SOC-10 PF6 (light intensity 2560 mJ/
cm2 min).
experiment shown in Figure 6. In both cases, the photopoly-
merizations take place essentially without an induction
period and the initial portions of the OP curves are nearly
identical. This is a reflection of the high reactivity of the
“kick-started” oxetane monomer system. However, thereafter,
the overall temperature versus time profiles of the two pho-
topolymerizations are different. As expected, SOC-10 SbF6
was a more reactive photoinitiator than SOC-10 PF6 produc-
ing a sharper and a higher exotherm. This is due to the fact
that HSbF6 generated on photolysis of SOC-10 SbF6 is a con-
siderably stronger acid and a more efficient polymerization
initiator than HPF6 produced by the photolysis of the corre-
sponding SOC-10 PF6 photoinitiator. When the integrated
curves were constructed from the two OP plots shown in
Figure 6, and the Rp values calculated from the slopes of the
initial portions of those curves, the rates are rather similar
(Rp SOC-10 SbF6 5 2.7; Rp SOC-10 PF6 5 2.0). It should be
pointed out that these results have considerable practical
influence the rate of photolysis of the photoinitiator such as
photoinitiator concentration, UV absorption characteristics,
quantum yields and acidity of the Brønsted acid generated
are expected to have a marked impact the overall rate and
also the induction period. In general, for a given photoinitia-
tor and “kick-started” oxetane monomer composition, those
factors that result in the highest rate of photoacid produc-
tion are the most beneficial in achieving the highest overall
rates. Several further studies were conducted to confirm
these conclusions.
A study of the effects of changes in the type of anion present
in the photoinitiator on an 8:1 molar DOX/LDO mixture is
depicted in Figure 6. The triarylsulfonium salt photoinitia-
tors, SOC-10 SbF6 and SOC-10 PF6 were compared at the
same molar concentrations in this study. It has been
observed that the quantum yields of photolysis of onium
salts is dependent on the structures of the organic cation
and not on the anion.1 For that reason, identical amounts of
the corresponding Brønsted acids, HSbF6 and HPF6 are pro-
duced per unit time at the irradiation intensity used in the
2
consequences. Photoinitiators bearing the PF6 anion are
2
less expensive than their counterparts bearing the SbF6
anion so there is an appreciable economic advantage for the
use of a photoinitiator containing the PF26 anion.
Figure 7 shows the effect of variations in the SOC-10 SbF6
photoinitiator concentration on the photopolymerization of
an 8:1 molar mixture of DOX and LDO. Reducing the photoi-
nitiator concentration from 1.0 to 0.25 wt %, produces only
a slight shift in the onset of the polymerization. Even at the
lowest photoinitiator concentration (0.25 wt %), the photo-
polymerization of the “kick-started mixture proceeds rapidly
and with nearly the same rapid exothermic behavior as
when a four-fold higher photoinitiator concentration was
employed.
As an alternative to optimizing the photoinitiator concentra-
tion in a given “kick-started” system to achieve the shortest
induction period and highest rate of polymerization, means
by which the effective quantum yield of the photoinitiator
decomposition can be increased were also explored. For
example, in Figure 8 is shown a comparative OP study in
FIGURE 7 Effect of variations in the SOC-10 SbF6 photoinitiator
concentration on the photopolymerization of an 8:1 molar mix-
ture of DOX and LDO (photoinitiator concentrations in wt %,
light intensity 1860 mJ/cm2 min)
which
a free radical photoinitiator (Irgacure 651, 2,2-
6
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
dimethoxy-2-phenylacetophenone) was used to induce the
decomposition of the diaryliodonium salt cationic photoini-
tiator. This effect is well understood.23–26 The photolysis of
both photoinitiators in this system proceed simultaneously
under broadband UV irradiation conditions. However, the
two photoinitiators possess different absorption characteris-
tics; the diaryliodonium salt has a band at 246 nm and Irga-
cure 651 has a long wavelength band at 340 nm. Combined,
the two photoinitiators absorb a larger portion of the
available emission from the broad band mercury arc light
source and channel that energy into photolysis of the
photoinitiators.
Unimolecular free radical photoinitiators such as 2,2-dime-
thoxy-2-phenylacetophenone undergo efficient photolysis by
an alpha-cleavage reaction as shown in eq 10 of Scheme 3 to
afford the benzoyl, 6, and dimethoxyphenylmethyl, 7, radi-
cals. The reduction of a diaryliodonium salt (eq 11) by 7
generates the dimethoxyphenylmethyl carbocation 8. There-
after, 8 initiates cationic ring-opening polymerization as
shown in eq 12 by direct electrophilic attack on the hetero-
cyclic monomer to given tertiary oxonium ion 9. The diphe-
nyliodine free radical, 10, irreversibly fragments to give a
phenyl radical and iodobenzene (eq 13). Scheme 3 consti-
tutes an independent parallel photochemical process by
which the diaryliodonium salt can be caused to generate
acid. This process, operating in tandem with the direct pho-
tolysis, results in an incremental increase in the amount of
strong acid formed during UV irradiation. This results, as
shown in Figure 8, in the observed decrease in the induction
period. Significant economic trade-offs can be achieved in
such “kick-started” oxetane systems through the replacement
of a portion of the relatively expensive diaryliodonium salt
photoinitiator with an inexpensive free radical photoinitiator.
FIGURE 9 Photopolymerization of DOX and an 8:1 molar mix-
ture of DOX and LDO in the presence of 3.0 wt % (0.55 mol %)
DAPSC1,C8 SbF6 as a photoinitiator.
former compounds proceeds by a reversible process to yield
a sulfonium ylid, 11, along with the Brønsted acid corre-
sponding to the anion of the starting sulfonium salt. It was
of interest to determine whether such sulfonium salts would
be efficient photoinitiators for use in “kick-started” oxetane
systems. As the results depicted in Figure 9 show, the
answer to this question is a definitive, yes. Employing 3.0 wt
% of S-n-octyl-S-methyl-S-phenacylsulfonium hexafluoroan-
timonate (DAPSC1,C8 SbF6) as the photoinitiator, the usual
long induction period (ꢁ33 sec) is observed for pure DOX.
However, when an 8:1 molar mixture of DOX and LDO were
used, the photopolymerization starts essentially without an
induction period. The photopolymerization of this latter mix-
ture is highly exothermic and the films obtained after UV
irradiation are colorless, clear and flexible. DAPSC1,C8 SbF6
is very soluble in the mixture of monomers.
Dialkylphenacylsulfonium salts, bearing anions derived from
strong Brønsted acids undergo photolysis by a different
mechanism than their diaryliodonium and triarylsulfonium
salt counterparts.27 As shown in eq 14, photolysis of these
CONCLUSIONS
Kinetic comparisons of the reactivity of “kick-started” oxe-
tanes with ERL as well as several other types of epoxide
monomers in cationic photopolymerizations have shown that
the former oxetane systems are considerably more reactive
than the latter epoxide ones. The results of the kinetic stud-
ies also indicate that in the photopolymerization of these
“kick-started” oxetane systems, an extremely favorable situa-
tion may be reached in which the rate of the photopolymeri-
zation is dependent on the rate of the photolytic generation
of the strong Brønsted acid from the photoinitiator. Based on
these results, it may be concluded that “kick-started” oxetane
systems may find uses in high speed curing of coatings,
printing inks, pressure sensitive adhesives, and in 3D print-
ing applications. They could replace epoxide monomers in
these as well as in many other commercial photocurable
applications.
SCHEME 3 Proposed mechanism of the photoinduced free rad-
ical decomposition of a diaryliodonium salt by 2,2-dimethoxy-
2-phenylacetophenone.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000
7

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
14 H. Sasaki, A. Kuriyama, In Proceedings of RadTech Asia
Radiation Curing Conference, Kuala Lumpur, Malaysia, August
26–24, 1999; pp. 263–268.
REFERENCES AND NOTES
1 A. G. Davies, Organic Peroxides; Butterworths: London, 1961.
2 S. Searles Jr, K. A. Pollart, F. Block, J. Am. Chem. Soc. 1957,
79(4), 952–956.
15 J. V. Crivello, U. Bulut, Design. Monomers Polym. 2005, 8
(6), 517–531.
3 D. B. Pattison, J. Am. Chem Soc. 1957, 79, 3455–3456.
16 M. Sangermano, G. Malucelli, R. Bongiovanni, Eur. Polym.
J. 2004, 40 (2), 353–358.
4 H. Sasaki, Photoinitiated cationic polymerization, Chapter 26,
In ACS Symp. Ser. 847; K. D. Belfield, J. V. Crivello, Eds.; Am.
Chem. Soc.: Washington, D.C., 2003; pp 296–305.
17 J. V. Crivello, J. Polym. Sci., Part A: Polym. Chem. 2014,
52(20), 2934–2946.
5 S. Inoue, T. Aida, In Ring Opening Polymerization, Chapter 4;
K. J. Ivin, T. Saegusa, Eds.; Elsevier: London, 1984; pp. 185–298.
18 J. V. Crivello, B. Falk, M. Jang, M. R. Zonca Jr, S. M.
Vallinas, RadTech Report, May/June 2004; pp. 36–43.
6 H. Sasaki, J. M. Rudzinski, T. Kakuchi, J. Polym. Sci., Part A:
Polym. Chem. 1995, 33, 1807–1816.
19 H. Sasaki, J. V. Crivello, J. Macromol. Sci., Pure Appl.
Chem. 1992, A29(10), 915–920; In Proceedings Radtech ’92 Con-
ference, Boston, MA, April 26, 1992; pp. 518–524.
7 H. Kato, H. Sasaki, Photoinitiated Cationic Polymerization,
Chapter 25; In ACS Symp. Ser. 847; K. D. Belfield, J. V.
Crivello, Eds.; Am. Chem. Soc.: Washington, D.C., 2003; pp.
285–295.
20 M. Matsumoto, H. Shigeta, N. Nakagawa, Eur. Patent
2,623,489,A1, August 7, 2013.
21 D. B. Pattison, U.S. Patent 2,924,607, February 9, 1960 (to
DuPont, de Nemours and Co.).
8 J. V. Crivello, J. L. Lee, J. Polym. Sci., Part A: Polym. Chem.
1989, 27, 3951–3968.
22 H. Sasaki, H. Photoinitiated Cationic Polymerization, Chap-
ter 26, In ACS Symp. Ser. 847; K. D. Belfield, J. V. Crivello,
Eds.; Am. Chem. Soc.: Washington, D.C., 2003, pp. 296–305.
9 S. R. Akhtar, J. V. Crivello, J. L. Lee, J. Org. Chem. 1990, 55,
4222–4225.
10 J. V. Crivello, S. Kong, Macromolecules 2000, 33 (3), 825–
832.
23 A. Ledwith, Macromol. Chem. Phys. 1979, 3, 348–358.
24 Y. Yagci, A. Ledwith, J. Polym. Sci., Part A: Polym. Chem.
1988, 26, 1911–1918.
11 B. Falk, S. M. Vallinas, J. V. Crivello, J. Polym. Sci., Part A:
Polym. Chem. 2003, 41(4), 579–596.
€
€
25 E. Klemm, H.-J. Flammersheim, R. Martin, H.-H. Horhold,
12 B. Falk, S. M. Vallinas, J. V. Crivello, Polym. Mater. Sci. Eng.
Prepr. 2003, 89, 279–280.
Angew. Makromol. Chem. 1985, 135, 131–138.
26 J. V. Crivello, S. Liu, Chem. Mater. 1998, 10 (11), 3724–
3731.
13 H. Sasaki, Toagosei, Co. Ltd. Web publication, www2.toago
sei.co.jp/develop/trend/No4/oxetanes.pdf. Last accessed on
November 27, 2014.
27 J. V. Crivello, J. L. Lee, Macromolecules 1983, 16, 864–870.
8
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 00, 000–000