Synthesis of cycloaliphatic substituted silane monomers and polysiloxanes for photocuring

Article abstract of DOI:10.1021/ma048998w

A synthetic scheme was developed to prepare cationically polymerizable methyl, cyclopentyl, and cyclohexyl substituted polysiloxanes. Initially, the desired cycloalkene and dichlorosilane were reacted at high pressure (approximately 250 psi) and high temp

Full text of DOI:10.1021/ma048998w

9402
Macromolecules 2004, 37, 9402-9417
Synthesis of Cycloaliphatic Substituted Silane Monomers and
Polysiloxanes for Photocuring
D. P. Dworak and M. D. Soucek*
Department of Polymer Engineering, The University of Akron, 250 S. Forge Street,
Akron, Ohio 44325-0301
Received May 20, 2004; Revised Manuscript Received September 22, 2004
ABSTRACT: A synthetic scheme was developed to prepare cationically polymerizable methyl, cyclopentyl,
and cyclohexyl substituted polysiloxanes. Initially, the desired cycloalkene and dichlorosilane were reacted
at high pressure (approximately 250 psi) and high temperature (120 °C) to yield the desired cycloaliphatic
dichlorosilane. The chlorosilane monomers underwent an oligomerization to produce cyclic oligomers of
low molecular weight (∼2000 g/mol). Polysiloxanes were produced through the acid-catalyzed ring-opening
polymerization of the cyclic oligomers to yield high molecular weight polysiloxanes (∼45 000 g/mol). The
polysiloxanes were then functionalized with a cycloaliphatic epoxy and alkoxysilane groups via
hydrosilylation. Monomers, oligomers, and polymers were characterized by 1H and 29Si NMR, FT-IR,
and electrospray ionization mass spectroscopy. The photoinduced curing kinetics and activation energies
were investigated using photodifferential scanning calorimetry. Differential scanning calorimetry was
used in order to observe any physical changes in the films that are brought about due to the variation of
the pendant groups. The cycloaliphatic substituents raised the glass transition temperature and affected
the curing kinetics when compared to a methyl substituted polysiloxane. The activation energies were
found to be 144.8 ( 8.1 kJ/mol for the methyl substituted and 111.0 ( 9.2 and 125.7 ( 8.5 kJ/mol for the
cyclopentyl and cyclohexyl substituted polysiloxanes.
Introduction
erization step. As a result, both stepsshydrolysis and
polymerizationsusually occur simultaneously during
the reaction.
Chlorosilanes are the basic building blocks of silicones
and polysiloxanes. Chlorosilanes are utilized in hydro-
silylation where the addition of a Si-H compound to a
multiple bond, often an alkene or alkyne, is the pathway
to synthesize a variety of unique polysiloxane systems.
Typically, platinum complexes are used as hydrosilyla-
tion catalysts due to activity at relatively low concentra-
tions. The most common platinum catalyst is chloro-
platinic acid (Speier’s catalyst) which is reduced to a
platinum(0) species in the presence of a silane or
siloxane. During these reactions an induction period was
observed, but this can be reduced by using a platinum-
(0) complex such as Karstedt’s catalyst. Other metal-
based catalysts can be used such as the rhodium-based
Wilkinson’s catalyst, although higher catalyst concen-
trations are usually required for efficiency.
Processes that involve synthesis of linear siloxane
polymers can be divided into two general classifications
according to the pathway in which the polymer chain
is formed: polymerization of difunctional silanes and
ring-opening polymerization of cyclic oligosiloxanes.
Hydrolytic polymerization involves the polymerization
of halosilanes (mainly chlorosilanes) through the incor-
poration of water and is often employed for the synthesis
of both linear siloxane polymers and cyclic siloxane
oligomers, the latter of which can be used further as
substrates in ring-opening polymerization. A polysilox-
ane chain is most often formed as a result of two types
of polymerization: homofunctional polymerization of a
difunctional silane such as silanediols or dichlorosilanes
and heterofunctional polymerization involving a silanol
and another functional group. A heterofunctional po-
lymerization approach is useful for a hydrolytic polym-
The process of ring-opening polymerization allows
greater control over molecular weight; thus, it is often
a preferred method for synthesis of high molecular
weight polymers. It can be performed via either anionic
or cationic routes and may be classified as either
thermodynamically or kinetically controlled. Most po-
lymerizations are thermodynamically driven; the silox-
ane bonds being identical in number and kind both in
a chain and in a ring, the net energy change is very
small. The increase in entropy that complements the
increased molecular freedom of the siloxane segments
on going from the cyclic structures to the linear chains
drives the polymerization reaction.1 The molecular
weight of the polymer at equilibrium during cyclosilox-
ane polymerizations is controlled by incorporating an
end group, which ensures the closure of the chain with
a neutral, nonreactive group.
Polysiloxanes possess a variety of applications, both
medical and nonmedical. Polysiloxanes and silsesqui-
oxanes have been functionalized with various reactive
(vinyl ethers, epoxy groups1-6) and nonreactive groups
(1-octene, substituted phenyl groups7,8) in order to
achieve the desired property. Applications of siloxanes
include high-performance elastomers, membranes, elec-
trical insulators, water repellants, antifoaming agents,
mold release agents, adhesives, and protective films.9,10
Furthermore, the ability to synthesize monomers that
increase the glass transition temperature could be
copolymerized with other polymers such as polyure-
thanes, polyimides, or other silicone monomers. In
previous work it was proposed to prepare poly(dimeth-
ylsiloxane-co-methylhydrosiloxane) (1) functionalized
with cycloaliphatic epoxides and alkoxysilanes via hy-
drosilylation chemistry.11 The principal silane mono-
* Corresponding author: Tel 1-330-972-2583; Fax 1-330-258-
2339; e-mail msoucek@uakron.edu.
10.1021/ma048998w CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/17/2004

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9403
an uncovered, hermetic, aluminum DSC pan. An empty pan
was used as a reference. The chamber of the DSC was purged
with nitrogen before the polymerization and was continued
throughout the reaction. The samples were photocured with
UV light (150 mW/cm2) for various exposure times (1, 5, and
15 s) and temperatures (-10, 25, and 60 °C). The heat flux as
a function of reaction time was monitored under isothermal
conditions, and the rate of polymerization was calculated.14,15
The heat of reaction (∆HR) used for the epoxy group was 23.13
kcal/mol.
The rate of propagation (Rp) is directly proportional to the
rate at which heat is released from the reaction. As a result,
the height of the DSC exotherm can be used in conjunction
with other sample information to quantify the rate of polym-
erization. The rate formula used in the analysis of the
photopolymerization data was
mers have been methyl and phenyl, and hitherto
cycloaliphatic silane monomers have not been reported
on account of steric hindrance of the cyclic alkenes.
The synthesis and functionalization of poly(dicyclo-
pentylsiloxane-co-cyclopentylhydrosiloxane), hydride ter-
minated (2), and poly(dicyclohexylsiloxane-co-cyclohex-
ylhydrosiloxane), hydride terminated (3), were performed
in a method comparable to the methyl and hydrogen
substituted polysiloxane. The synthetic route for the
cycloaliphatic substituted differs from that of methyl
substituted in that the cyclic species needed to be
synthesized via hydrolytic polymerization of cycloaliphat-
ic substituted silanes. The monomers were characterized
1
using H NMR, 29Si NMR, FT-IR, and mass spectros-
copy. From the monomers, oligomers, and polysiloxanes
homopolymers were prepared. The synthesis of these
monomers makes it possible to examine the unique
properties that large bulky pendant groups can induce
on high molecular weight polysiloxane chains and to
explore the various applications for these polymers.
Rp ) ((Q/s)M)/(n∆HRm)
(3)
where (Q/s) is the heat flow per second released during the
reaction in J/s, M is the molar mass of the reacting species, n
is the average number of epoxy groups per polymer chain, and
m is the mass of the sample.
Experimental Section
Synthesis of Poly(dimethylsiloxane-co-methylhydro-
siloxane), Hydride Terminated. To a three-neck round-
bottom flask equipped with a reflux condenser and nitrogen
inlet/outlet was added octamethylcyclotetrasiloxane (90.00 g,
0.30 mol), 1,3,5,7-tetramethylcyclotetrasiloxane (5.33 g, 22.1
mmol), 1,1,3,3-tetramethyldisiloxane (0.67 g, 5.3 mmol), and
Amberlyst 15 (20 wt %) and stirred at 70 °C, under nitrogen,
for 15 h. The viscous solution was then filtered to obtain poly-
(dimethylsiloxane-co-methylhydrosiloxane), hydride termi-
nated (1), of various molecular weight ranges. Vacuum filtra-
tion was performed (<1 mmHg) in order to remove low
molecular weight oligomers and unreacted starting materials.
Weight-average molecular weight was obtained from gel
permeation chromatography (GPC) analysis; Mn ) 45 000, PDI
) 1.66. Polymer characterization and Si-H functionality were
confirmed/analyzed through 29Si NMR, 1H NMR, FT-IR analy-
sis, and titration. Polysiloxanes 2 and 3 were produced in the
same manner.
Materials. Octamethylcyclotetrasiloxane, 1,3,5,7-tetra-
methylcyclotetrasiloxane, 1,1,3,3-tetramethyldisiloxane, di-
chlorosilane, and vinyltriethoxysilane were purchased from
Gelest, Inc., and used as supplied. Wilkinson’s catalyst (chlo-
rotris(triphenylphosphine)rhodium(I), 99.99%), Karstedt’s cata-
lyst (platinum(0)-1,3-divinyl-1,1,3,3-tetramethyldisiloxane com-
plex, 3% w/w solution in xylenes), cyclopentene, cyclohexene,
Amberlyst 15 ion-exchange resin, and 4-vinyl-1-cyclohexene
1,2-epoxide were purchased from Aldrich and used as supplied.
Toluene, supplied by Aldrich Chemical Co., was distilled in
order to eliminate any impurities. Irgacure 250 was supplied
by Ciba Specialty Chemicals and used as received. Air-
sensitive materials were transferred and weighed in a drybox
under argon.
Instruments. Proton NMR spectra were obtained from a
Gemini-300 spectrometer (Varian), while silicon NMR spectra
were recorded on a Gemini-400 spectrometer (Varian). All
NMR samples were prepared in CDCl3 and recorded at 20 °C.
Chemical shifts are given relative to a TMS internal standard.
FT-IR spectra were obtained on a Mattson Genesis Series
FTIR, and a Waters system was used for GPC analysis. Mass
spectroscopy was performed on a Saturn 2200 (Varian) in EI
mode with an ion trap readout.
Synthesis of Cycloaliphatic Dichlorosilane. A dry,
sealed, and evacuated stainless steel bomb, cooled via a dry
ice/acetone bath, was charged with chilled cycloalkene (5 g,
∼30 mmol) and Wilkinson’s catalyst (0.15 g, 0.16 mmol) and
purged with nitrogen. In a chilled (<-10 °C) calibrated tube
dichlorosilane (5 mL, 0.06 mol) was condensed and then
distilled into the bomb through the inlet valve via a cannula.
The inlet valve was sealed and the bomb was then allowed to
warm to room temperature and then heated for 24 h at 120
°C by means of an oil bath. The bomb was then allowed to
cool, and the reaction produced a clear, light yellow liquid.
After distillation, any unreacted cycloalkene and side products
were removed via vacuum (2-3 mmHg) to yield pure cy-
cloaliphatic dichlorosilane (∼88% yield). Product characteriza-
Functional Group Analysis. The Si-H bond is polarized
depending to some degree on the substituents of the silicon.12
The reactivity of the Si-H bond makes it possible to analyze
this group with qualitative or quantitative chemical tests. The
Si-H was titrated via reduction of a mercury(II) salt:
-Si-H + 2HgCl2 f -Si-Cl + Hg2Cl2 + HCl
(1)
which can be titrated with a base to find the % Si-H in a given
sample.13
1
tion was performed by 29Si NMR, H NMR, FT-IR, and mass
A mercuric chloride solution (4% w/v in 1:1 chloroform-
methanol) was pipetted (20 mL) into an Erlenmeyer flask. The
sample to be titrated was added and agitated before adding
the calcium chloride solution (15 mL, saturated solution in
methanol). Phenolphthalein indicator was added (15 drops)
after 5-6 min, and the solution was titrated with 0.1 N
alcoholic potassium hydroxide. Blanks are titrated in the same
manner before and after the analysis. The calculation for %
H is shown in eq 2:
spectroscopy.
General Synthesis of Cyclic Oligomers of Poly(cy-
cloaliphatic hydrosiloxane). To a three-neck round-bottom
flask, equipped with a reflux condenser, nitrogen inlet/outlet,
and dropping funnel, was added saturated aqueous sodium
bicarbonate (10 mL) and diethyl ether (5 mL). A solution of
cycloaliphatic dichlorosilane (4.43 g, ∼0.03 mol) in ethyl ether
(5 mL) was then added dropwise via the dropping funnel, and
the solution was allowed to stir for several minutes at room
temperature. The ether layer was separated and passed
through a filter, and any remaining traces of ether were
removed via vacuum distillation (3-5 mmHg) to yield a clear,
viscous oil. Weight-average molecular weight was obtained for
both the cyclopentyl and cyclohexyl substituted cyclic oligomers
from gel permeation chromatography. The poly(cyclopentyl-
hydrosiloxane) oligomers had a Mn ) 1800 and a PDI ) 2.44.
The poly(cyclohexylhydrosiloxane) oligomers had a Mn ) 2230
% H ) [(V1 - V2)](NKOH)(1.008/2000)(100)/sample wt (g)
(2)
where V1 is the end point, V2 is the averaged blanks, and N is
the normality of the basic titrant.
Photopolymerization Procedure. On average, 2-3 mg
of sample (polymer and 3% w/w photoinitiator) was placed in

9404 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
and a PDI ) 2.53. Oligomer characterization was performed
1
via 29Si NMR, H NMR, FT-IR, and mass spectroscopy.
Synthesis of Cyclic Oligomers of Poly(dicycloaliphatic
siloxane). To a single-neck round-bottom flask equipped with
a reflux condenser was added the cyclic oligomers of the
desired poly(cycloaliphatic hydrosiloxane) (5 g), the preferred
cycloalkene (15 g), and Karstedt’s catalyst (0.1 mL, 0.22 mmol).
The reaction was held at 110 °C in an oil bath and magneti-
cally stirred. The disappearance of the Si-H functionality was
monitored through FT-IR, and the disappearance of the peak
at ∼2160 cm-1 indicated that the reaction was complete (>48
h). Any unreacted cycloalkenes were removed under vacuum
(3-1 mmHg) to yield a clear, viscous oil. Products were
characterized by 29Si NMR, 1H NMR, FT-IR, and mass
spectroscopy.
Cycloaliphatic Epoxide and Alkoxysilane Function-
alization of Prepared Poly(dialkyllsiloxane-co-alkyl-
hydrosiloxane), Hydride Terminated Polymers. To a
three-neck round-bottom flask, equipped with a reflux con-
denser and nitrogen inlet/outlet, was added 1, 2, or 3 (30 g),
4-vinyl-1-cyclohexene diepoxide (20 g, 0.18 mol), vinyl tri-
ethoxysilane (2 g, 0.01 mol), and Wilkinson’s catalyst (0.004
g, 4.3 µmol). Dry, distilled toluene (30 g) was added via a
cannula. The reaction was held at 75 °C in an oil bath and
mechanically stirred under nitrogen. The disappearance of the
Si-H functionality was monitored through FT-IR, and the
disappearance of the peak at ∼2160 cm-1 indicated that the
reaction was complete. Any solvent and unreacted starting
materials were removed under vacuum (3-5 mmHg). Cy-
cloaliphatic epoxide and alkoxysilane functionalization was
confirmed/analyzed through 1H NMR, FT-IR analysis, and
titration.16
Results and Discussion
Figure 1. Synthesis of cycloaliphatic epoxide/alkoxysilane
functionalized poly(dimethylsiloxane-co-methylhydrosiloxane).
The objective of this study was to prepare a UV-
curable polysiloxane with cyclopentyl and cyclohexyl
groups. The route used in the preparation of the
cycloaliphatic substitution was taken due the difficulties
of reacting the internal alkene with the available silane
group. The initial hydrosilylation between dichlorosilane
and the cycloalkene was performed in order to control
the amount of silane functionality along the polysiloxane
backbone. The reaction of the mono- and disubstituted
cycloaliphatic substituted cyclic oligomers is similar to
that of the methyl substituted polysiloxane (Figure 1).
The control over the ratio of mono- and disubstituted
cyclic oligomers allows management over the silane
functionality present along the polymer backbone. In
these experiments, all mono- and disubstituted cyclic
oligomer ratios were 1:18.
Attempts at the synthesis of the dicycloaliphatic
dichlorosilanes proved exceptionally difficult even under
extreme pressure (300 psi) and temperatures (250 °C).
Hydrosilylation accelerators, such as benzaldehyde and
acetone, and higher catalyst concentrations were em-
ployed with no considerable results.
Synthesis and Characterization. A synthetic dia-
gram of the functionalization of poly(dimethylsiloxane-
co-methylhydrosiloxane), hydride terminated, is pre-
sented in Figure 1. The pendant alkoxysilane aids in
miscibility during formulation and provides a site for
interaction with the metal/silicon oxo cluster, while the
cycloaliphatic epoxide provides a cross-linking site for
cationic UV induced cure. The overall structure of 1 will
be a block copolymer composed of mostly “D” units
(R-Si-R) and the desired functionalities.
A diagram of the synthesis of 2 is presented in Figure
2, and functionalization of 2 and 3 is performed in the
same manner as for 1. The cycloaliphatic substitution
along the polysiloxane backbone was chosen for raising
Figure 2. Synthesis of hydride functionalized poly(dicy-
cloaliphatic siloxane-co-cycloaliphatic hydrosiloxane).
the glass transition of the polysiloxane, while not
contributing to UV absorption and ultimately yellowing.

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9405
Figure 3. FT-IR spectra of poly(dimethylsiloxane-co-methylhydrosiloxane), hydride terminated.
Figure 4. Proton NMR of poly(dimethylsiloxane-co-methylhydrosiloxane), hydride terminated.
1
The synthesis of 1 was verified through FT-IR (Figure
3) and showed that the Si-H functionality is present
due to the strong peak at 2160 cm-1, while the two
peaks 1090 and 1110 cm-1 are indicative of high
molecular weight polysiloxanes. The indication of cyclic
species would be evident in the region of 1000 cm-1 with
the band being strong and broad and spanning 50-100
cm-1 across. The indication of two bands in this region
is strong confirmation of high molecular weight species.
However, cyclic species higher than 50 units long can
still show similar peaks and present misleading infor-
mation. As a result, GPC analysis was used to confirm
that no traces of cyclic species were present. The GPC
chromatogram showed the typical bell curve for high
molecular weight polymer, and if cyclic species were
present, then two more peaks would have eluted before
the main peak, which are evidence of cyclic species. The
% Si-H analysis for 1 yielded 15.3 ( 0.8%.
Further analysis via H NMR (Figure 4) showed the
strong, characteristic peak for Si-Me at δ 0.07 ppm and
the Si-H functionality at δ 4.68 ppm. The peak at δ
0.07 ppm is split due to an adjacent Me-Si-H atom.
Two smaller peaks were observed at δ 1.22 and δ 1.57
ppm, which are a result of magnetic inequalities due to
the different stereoconfigurations of the substituents
along the polymer backbone. The magnetic nonequiva-
lence of the CH3 protons is a result of their different
environment due to fact that one methyl group has two
other methyl groups as neighbors in the cis-cis position,
whereas these others are surrounded by one methyl and
one hydrogen in a cis-trans position.
The silicon NMR (J ) 200) of 1 (Figure 5) shows to
main groups of signals corresponding to the D and D′
(Me-Si-H) units in the difunctional siloxy unit region.
A typical copolymer whose molar ratio is 1:1 and/or have
random units show more symmetric peak patterns.

9406 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
Figure 5. Silicon NMR of poly(dimethylsiloxane-co-methylhydrosiloxane), hydride terminated.
Figure 6. FT-IR spectra of cycloaliphatic epoxide/alkoxysilane functionalization of poly(dimethylsiloxane-co-methylhydrosiloxane),
hydride terminated.
However, when the molar ratio is no longer equal (14:1
in this case) and/or the copolymer is a blocky micro-
structure, then the intensities in the NMR patterns
deviate from the symmetric case. Furthermore, the
observance of a quartet can occur from the nonadditivity
of the chemical shifts resulting from different arrange-
ments of the neighboring units.
The small downfield peak at ∼δ -8 ppm is represen-
tative of a hydrogen substituted “M” (R3-Si-) unit.
Trimethyl substituted M units appear in the δ 5-10
ppm region of the spectra, which is further evidence that
the polysiloxane chain is hydrogen terminated. When
comparing the D unit silicon atoms to an M unit one,
the D unit silicon is bonded to an additional oxygen
atom. The downfield shift of the M unit occurs due to
the lack of an oxygen atom, which deshields the Si
nuclei. The cluster of peaks near δ -21 ppm are
representative of various D units. The inset shows peaks
between δ -20 and δ -21 ppm, which are representa-
tive of D units that are adjacent to a D′ unit. The

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9407
Figure 7. Proton NMR spectra of cycloaliphatic epoxide/alkoxysilane functionalization of poly(dimethylsiloxane-co-methyl-
hydrosiloxane), hydride terminated.
Table 1. Calculated and Actual Values of Si-H
nonequivalence of the of the silicon atoms causes a slight
Transmittance Peaks
downfield shift. The peaks at δ -22 ppm symbolize
structure
calculated (cm-1
)
actual (cm-1)
repeating D units in a linear chain. The various peaks
are a result of the different molecular weight chains in
the sample. The upfield peaks at ∼δ -38 ppm are the
D′ units within the polysiloxane chain, and the same
trend with the D units is observed with the D′ units,
with the repeating units being slightly upfield of the
different adjacent silicon atom peaks. In addition, the
starting material 1,3,5,7-tetramethylcyclotetrasiloxane
has reacted completely due to the fact that it would
appear near δ -32 ppm in the spectrum.
cyclopentyldichlorosilane
cyclohexyldichlorosilane
2206
2197
2202
2200
representative of the substituent due to the shielding
effects of the silicon atom.18 However, this pattern can
be affected by the inductive and/or shielding effects of
the other substituents on the silicon atom. Closer
examination of the Si-H peak reveals a multiplet,
which could be the result of the sample reacting with
residual water left in the CDCl3 and producing oligo-
mers of 3-6 units long. The presence of the small pair
of satellite lines near the main resonance of the Si-H
peak is a result of the 29Si isotope. The locations of the
Si-H peaks between the cyclopentyl and cyclohexyl
group differ by approximately δ 0.1 ppm but show that
only subtle changes in the substituents can affect the
position of the Si-H peak.
Silicon NMR was also used to analyze the reaction
product (Figure 10). Figure 13 shows that several side
products are formed in addition to the desired product.
The formation of the disilane compounds (tetrachloro-
disilane and dicyclopentyltertrachlorosilane) is of par-
ticular interest in that the catalyst used is not only
selective toward hydrosilylation through alkenes but
also can undergo addition reactions to form disilane
compounds. Preliminary data show that bulky substit-
uents tend to hinder the formation of disilane com-
pounds, while the smaller groups yield more.
Functionalization of 1 with cycloaliphatic epoxides
and alkoxysilanes was analyzed with FT-IR (Figure 6)
and shows clear evidence of the epoxy ring. However,
the Si-O-Si bands at 1000 cm-1 take precedence and
mask the alkoxysilane functionality. As a result, proton
NMR was used for confirmation of the alkoxysilane
functionality (Figure 7). The triplet at δ 3.8 ppm
represents the -CH2- protons of the alkoxysilane
group.
The synthesis of the cycloaliphatic dichlorosilanes was
analyzed through FT-IR, proton NMR, and silicon NMR.
The FT-IR spectra of the cycloaliphatic dichlorosilanes
(Figure 8) show the presence of the Si-H peak at ∼2200
cm-1 and the existence of the Si-Cl2 functionality at
∼500 cm-1. The Si-H transmission peak location can
be a good indication of the inductive effects of the other
substituents on the silicon atom.17
The silane peak position can be predicted accurately
if all of the substituents are known. By adding the
wavelength values for each of the substituents (-Cl,
-Cl, and -C5H9/C6H11), the location the Si-H peak
value can be found (Table 1).12
The proton NMR spectra of the cycloaliphatic di-
chlorosilanes (Figure 9) show three distinct peaks. The
upfield peaks are representative of the cycloaliphatic
substituent, while the downfield peak is the Si-H
proton. The Si-alkyl groups usually display the ex-
pected shift patterns, with the patterns being fairly
After distillation the silicon NMR (Figure 11) reveals
the presence of the desired cycloaliphatic substituted
silane. The difference between the two peaks is only
0.276 ppm, showing that one less carbon atom has
enough “deshielding” character to cause an upfield shift
in the peak signal.
The mechanistic fragmentation of R4Si compounds,
where R is alkyl, aryl, hydride, and others (symmetrical
or unsymmetrical), has been studied.19 Figure 12 shows

9408 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
Figure 8. FT-IR spectra of (a) cyclopentyldichlorosilane and (b) cyclohexyldichlorosilane.
the mass spectra for both cyclopentyldichlorosilane and
cyclohexyldichlorosilane and shows the radical cleavage
of each of the four substitutes. For silanes containing
the substituents present on the above samples, the
larger alkyl substituents tend to cleave first, which is
observed due to the abundance of the cycloaliphatic ring
in both of the spectra. In addition, when the breakage
of the C-C bond becomes less and less probable, as in
the case with cyclohydrocarbons and short alkyl chains,
another process competes. The process is the loss of HCl
from the molecular ion, leaving another odd electron ion.
Figure 12 contains evidence of this loss (m/z 133 for
cyclopentyldichlorosilane) and (m/z 147 for cyclohexyl-
dichlorosilane). The presence of the molecular ion,
predictable radical cleavages, and a distribution pattern
which allows recognizable isotropic peaks makes the
interpretation for this class of compounds straight-
forward.
seriously affect the molecular weight and overall linear-
ity of the polymerized product unless the acid is
neutralized from the system. The substituents are
usually resistant toward the HCl that is given off as a
byproduct, such as dichlorodimethylsilane. However, if
a Si-H group is present, the HCl released will react
with it
Si-H + HCl f Si-Cl + H2
(4)
and form an undesired Si-Cl, which can undergo
hydrolysis and rather than a linear polysiloxane chain
a substituted silsesquioxane is formed. Therefore, a
saturated aqueous basic solution is used to neutralize
the liberated acid.
The products of the hydrolytic polymerization of the
cycloaliphatic dichlorosilanes were analyzed via FT-IR
(Figure 13). The strength of the Si-O-Si band indicates
that the rings are not of sufficient size (<20) due to the
fact that the characteristic antisymmetric stretch of Si-
The acid liberating polymerization of dichlorosilanes
is an equilibrium process; the reverse reaction may

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9409
Figure 9. Proton NMR spectra of (a) cyclopentyldichlorosilane and (b) cyclohexyldichlorosilane.
Figure 10. Silicon NMR of dicyclopentyldichlorosilane reaction before distillation.

9410 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
Figure 11. Silicon NMR spectra of (a) cyclopentyldichlorosilane and (b) cyclohexyldichlorosilane.
O-Si at ∼1100 cm-1 peak is not present, as seen in
Figure 8. What is also important to note is the disap-
pearance of the Cl-Si-Cl peaks near the 500 cm-1
region, indicating that all of the material had reacted
into polysiloxane chains. Another interesting note is the
nonexistent Si-OH peak, typically found near the 3700
cm-1 region. This lack of this peak is evidence that cyclic
species are present and that no linear chains terminated
with a terminal Si-OH group were produced. Finally,
the Si-H peak was still present, showing that the HCl
produced during the reaction was neutralized by the
base and did not affect the functionality.
The proton NMR spectra of the cyclic oligomers show
the same trend as seen in the NMR spectra of 1 (Figure
4). Since the substituents along the polysiloxane chains
are atactic, their environments are different, which
causes the protons to possess magnetic nonequivalence.
When the spectra in Figure 14 are compared with those
in Figure 9, it is apparent that the cis/trans configura-
tions have a dramatic effect on the NMR spectra. When
studying the Si-H proton peak at ∼δ 4.5 ppm, it is
evident that more than one signal is present and the
splitting appears to be more predominant in longer
chains. This is, again, due to the atactic configurations
of the substituents and the hydrogen atom being either
surrounded by two hydrogens, one hydrogen and one
cycloaliphatic ring, or two cycloaliphatic rings. The same
reasoning is behind the multitude of peaks in the region
for the cycloaliphatic protons; however, the effect seems
to be more prominent with the Si-H protons.
The silicon NMR (J ) 200) of the cyclic oligomers
(Figure 15) shows evidence of cyclic species. The down-
field peaks represent the cyclic D′ units of low molecular
weight oligomers, while the upfield peaks are charac-
teristic of larger cyclic D′ units. The peaks present at
-30 ppm are indicative of R-Si-H siloxane units, while
the peaks near δ -20 ppm represent smaller cyclic
siloxane structures.20,21 The ring strain that the small
cyclic siloxanes possess deshields the 29Si nucleus,
resulting in shifts to higher frequencies. The multiple
peaks, again, come from the number of microenviron-
ments present along the oligomer chain and the various
sized cyclic structures. Figure 15 shows that no Si-OH
groups exist and also that all the Si-Cl groups have
reacted, which would be the possible end groups on
linear chains.
Calculations show that the average size of the cyclic
structures is 15 units.13 The small size of the cyclic

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9411
Figure 12. Mass spectroscopy analysis of (a) cyclopentyldichlorosilane and (b) cyclohexyldichlorosilane.
chains could be a result of the dilution of the cy-
cloaliphatic dichlorosilanes along with the dropwise
method of adding the dissolved silanes into the organic
phase. It is interesting to note that both oligomerization
reactions, regardless of the substituent, produced ap-
proximately 10% of small cyclic species of about 3-4
siloxane units.
Matrix-assisted laser desorption ionization-time-of-
flight mass measurement was attempted to determine
the molecular weight and confirm the repeat units.
Obtaining this information from the cyclic oligomers
proved difficult in that the bulky cyclic groups shield
the oxygen atoms from cation attachment. In addition,
cyclic species do not readily ionize as a result of cyclic
structure hindering coordination of the large sodium ion.
A smaller ion (lithium) was used in the likelihood that
it was small enough to coordinate with the oxygen
atoms, but this attempt also did not yield appreciable
results. Quadrupole time-of-flight mass spectroscopy
was also attempted but generated the same noisy data.
Various ratios of matrix, cation, and sample were tested
with no satisfactory results. The highest molecular
weight obtained from GPC was ∼4000, which would
have a ring size of approximately 37 units.
The proton NMR spectra of the results from the
hydrosilylation of 2 and 3 with the desired cycloalkene
to produce the cyclic oligomers of poly(dicycloaliphatic
siloxane) (Figure 16) show the disappearance of the
Si-H functionality (∼δ 4.5 ppm) and the strong pres-
ence of the cycloaliphatic character. The reaction was
sluggish due to the bulky substituents and that the
alkene functionality undergoing hydrosilylation was an
internal alkene as opposed to the faster reacting ter-
minal alkene. The cyclic structure of the polysiloxanes

9412 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
Figure 13. FT-IR spectra of cyclic oligomers: (a) polycyclopentylhydrosiloxane; (b) polycyclohexylhydrosiloxane.
also played a role in the lengthy rate of hydrosilylation.
The FT-IR spectra also confirmed the disappearance of
the Si-H functionality.
end-capper the chains will terminate by pure chance
and will extended the lifetime of the reaction. The %
Si-H analyses for 2 and 3 were 11.5 ( 1.1% and 10.8
( 0.9%. Gel permeation chromatography showed that
both 2 and 3 had average molecular weights of ∼35 000
g/mol.
Photo and Differential Scanning Calorimetry.
Photoinitiated cationic polymerization of functionalized
polysiloxanes yields highly cross-linked networks. In
addition, it is important to understand the relationship
between the polymer structure and reaction conditions
with the properties of the networks formed. Photodif-
Once the cyclic oligomers of the poly(dicycloaliphatic
siloxane)s were prepared, they could be combined with
the cyclic oligomers of the poly(cycloaliphatic hydro-
siloxane)s to yield poly(dicycloaliphatic siloxane-co-
cycloaliphatic hydrosiloxane), hydride terminated (2 and
3). The procedure to produce 2 and 3 is done in the same
matter as it is to produce 1, which is employing A-15
ion-exchange resin and an end-capper; which is used
in order to control the molecular weight. Without the

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9413
Figure 14. Proton NMR spectra of cyclic oligomers: (a) polycyclopentylhydrosiloxane; (b) polycyclohexylhydrosiloxane.
ferential scanning calorimetry (PDSC) makes it possible
to investigate the polymerization behavior, and espe-
cially the kinetics of photopolymerization reactions, due
to their highly exothermic reaction characteristic upon
exposure to UV light. As a result, the reaction rate may
be measured by observing the rate at which the heat is
released from the sample subject to polymerization.
Profiles of reaction heat releases vs time can be utilized
to characterize the reaction kinetics.22-24 The kinetics
vary from system to system and are generally complex;
therefore, a generalized kinetic expression is not avail-
able for cationic polymerizations.25 This is due to the
reactivity of the carbocationic center being heavily
dependent on the proximity of the counterion.
steady-state approximation is invalid for these types of
polymerizations due to the active centers not undergoing
combination, as seen in free radical polymerizations. As
a result, the rates of initiation and termination are not
equal, which require non-steady-state analysis. The
understanding of the kinetics of cationic polymerizations
is valuable due to the increasing demand for fast curing,
low VOC films. The PDSC profiles obtained present
significant information about the curing kinetics of
cationic photopolymerizations.
Reactions were carried out at various temperatures
(-10, 25, and 60 °C) in order to establish an overall
activation energy for the polymerization. Reactions were
also studied with various exposure times (1, 5, and 15
s) in order to observe any changes in the rate of
polymerization. Figure 17 is a typical exotherm obtained
from the photoinduced polymerizations.
In addition, a number of propagating species may be
identified during the polymerization such as ion pairs,
solvated ions, and/or aggregates.12 Finally, the pseudo-

9414 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
Figure 15. Silicon NMR spectra of cyclic oligomers: (a) polycyclopentylhydrosiloxane; (b) polycyclohexylhydrosiloxane.
Figure 16. Proton NMR spectra of cyclic oligomers: (a) polydicyclopentylsiloxane; (b) polydicyclohexylsiloxane.

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9415
Figure 17. Exotherm of the cationic polymerization of 1 at 25 °C for 15 s.
Table 2. Effect of Substituent and Temperature on the Rate of Polymerization
av mol wt (Mn)
exposure
time (s)
heat flow per
second (J/s)
activation
substituent
methyl
(g/mol)
temp (°C)
Rp (s-1
)
conv (%)
energy (kJ/mol)
45 500
34 530
38 290
-10
25
60
-10
25
60
-10
25
60
5
5
5
5
5
5
5
5
5
9.456
14.900
19.646
8.223
5.892
11.723
8.113
0.0093
0.0110
0.0125
0.0025
0.0055
0.0057
0.0054
0.0047
0.0068
99.6
99.5
99.7
90.4
97.2
97.4
91.2
96.7
98.2
144.8 ( 8.1
111.0 ( 9.2
125.7 ( 8.5
cyclopentyl
cyclohexyl
5.932
16.890
The degree of cure, or conversion, can be estimated
from the ratio of the amount of heat evolved from the
partial conversion after time t at a specific temperature
(Ht) to the total heat evolved from the reaction, ∆Hp:
ature. This is demonstrated from the exotherms exhibit-
ing a larger integrated heat as the temperature was
increased.
Increasing the size of the substituent has a dramatic
effect on the rate of polymerization and total conversion
(Table 2). Figure 18 also shows that as the size of the
pendant group is increased, the glass transition tem-
perature increases. The rates of polymerization were
decreased by an overall average of 50% when comparing
a methyl substituted to a cycloaliphatic substituted
polysiloxane. This could be attributed to the large bulky
substituents hindering molecular motion during the
reaction and preventing the active species from further
polymerization. As expected, the longer exposure to UV
light yields a higher conversion due to the production
of more active species. The percent conversion calcula-
tion is taken from the moment the exotherm begins to
the when the slope of the thermogram is equal to zero.
The effects of “postcure” are not taken into account in
the final calculation. The consequence of postcure would
undoubtedly increase the final percent conversion value.
R ) Ht/∆Hp
(5)
If eq 5 is a function of conversion, but not temperature,
as in photoinduced experiments, the activation energy,
E, can be obtained by plotting ln[(1/∆Hp)(dH0/dt)] vs
(1/T), where (dH0/dt) is the heat of polymerization at
the maximum peak of the exotherm. Calculating the
slope of this plot will yield the activation energy for the
polymerization (Table 2). Photo-DSC experiments record
the total heat of polymerization; therefore, the activation
energy is representative of an overall activation energy
which includes initiation, activation, and termination:
ER ) EP + EI - ET
(6)
In order for the above equation to be valid, the produc-
tion of active centers must remain throughout the
reaction. Previous measurements26 analyzed the kinetic
activity of photosensitizers and showed that the photo-
sensitizers were not completely consumed until after the
reaction had progressed. Indicating that the active
center concentration had continued to increase after the
peak maximum, which was were the reaction rate was
measured. As a result, eq 6 can be used to signify the
overall activation energy for the photoinduced polym-
erization reaction. As anticipated, the reaction rate and
total conversion increased with an increasing temper-
The activation energies for the cycloaliphatic substi-
tuted polysiloxanes are lower than the methyl substi-
tuted polysiloxane value. The higher average molecular
weight (Mn) of the methyl substituted polysiloxane
reduces the mobility of the epoxy groups, which leads
to a higher activation energy.27,28 The dependence on
activation energy with viscosity could also account for
the slight difference in activation energy in the cy-
cloaliphatic substituted polysiloxanes.
Figure 18 illustrates that the glass transition tem-
perature can be tailored by being able to add on the

9416 Dworak and Soucek
Macromolecules, Vol. 37, No. 25, 2004
Figure 18. Overlay of Tg analysis of cured coating with 3% photoinitiator.
desired pendant group through hydrosilylation. The
pendant group could be chosen from a number of alkene
functionalized systems.
nal alkenes was possible at high pressure and temper-
atures but became sluggish when the silane function-
ality became sterically hindered as seen in the hydro-
silylation between the cyclic oligomers and the cyclo-
alkene. In addition, the ability to maintain the silane
functionality throughout the hydrolytic polymerization
makes it possible for additional hydrosilylation between
the polysiloxane chain and desired alkene. Analysis of
the hydrolytic polymerization reaction showed the pro-
duction of a wide range of cyclic oligomers, while the
use of the sulfonated ion-exchange resin made the
production of high molecular weight polysiloxanes par-
ticularly straightforward. Photo-DSC exotherms showed
that the bulky pendant groups affect the kinetics by
lowering both the rate of polymerization and overall
conversion, when compared to the methyl substituted
polysiloxane. This is due to the bulky pendant groups
interfering with molecular motion. As expected, the
increase in UV exposure time and reaction temperature
resulted in an increase in the reaction rate for all of the
substituted polysiloxanes. This increase yielded higher
final conversions as revealed from the total heat of the
reaction.
The ability to synthesize polysiloxanes with cy-
cloaliphatic pendant groups will be beneficial in that it
provides more monomers to be considered for new
materials, membranes, films, or other potential applica-
tions. The capability to modify and/or tailor the pendant
groups could also solve miscibility and grafting issues
by making the groups similar in structure and/or
polarity to a specific type of solvent. Furthermore,
tailoring the polysiloxanes with cycloaliphatic epoxides
and alkoxysilane groups to UV curable hybrid films
could be used in the application toward membranes or
low coefficient of friction coatings.29-31 The ability to
adjust the glass transition temperature through the
variation of the pendant groups will also be beneficial
for other specialized applications such as lubricants for
fibers, wetting agents for polyurethane foams, and
temperature-sensitive coagulating agents for latexes.32
The need to understand and optimize curing condi-
tions for inorganic/organic materials is critical for the
in situ formation of nanocomposite materials.33 The rise
in cationic curing chemistry is increasing due to the
unique properties of ring-opening polymerization, which
include inhibition of oxygen and the improved adhesion
over free-radical UV-cured films. The PDSC experi-
ments were used to determine the rate of polymerization
and activation energies for the synthesized polysilox-
anes in order to determine what effect the larger
substituents have on the kinetics of polymerization in
addition to temperature and UV exposure time. The
knowledge of cure kinetics and the ability to customize
polysiloxanes unlocks new potential in the development
of innovative polymers. In a later study, siloxane
membranes are studied, and the effect on cycloaliphatic
groups on permeability will be reported.
Acknowledgment. A special thanks to Kathryn
Wollyung for her mass spectroscopy assistance and to
Dr. Venkat Dudipala for his silicon NMR contribution.
References and Notes
(1) Smith, A. L. The Analytical Chemistry of Silicones; Wiley:
New York, 1991.
(2) Mowers, W. A.; Rajarman, S. K.; Liu, S.; Crivello, J. V. New
York State Center for Polymer Synthesis, Department of
Chemistry, Rensselaer Polytechnic Institute, Technical Con-
ference Proceedings, 2000, Baltimore, MD, Apr 9-12, 45.
(3) Crivello, J. V.; Lee, J. J. Polym. Sci., Part A: Polym. Chem.
Ed. 1990, 28, 479.
(4) Oishi, M.; Kawakami, Y. Macromolecules 2000, 33, 1960.
(5) Belfield, K. D.; Chinna, C.; Najjar, O. Macromolecules 1998,
31, 2198.
(6) Chojnowski, J.; Cypryk, M.; Fortuniak, W.; Scibiorek, M.;
Ro´zga-Wijas, K. Macromolecules 2003, 36, 3890.
(7) Crivello, J. V.; Malik, R. Silicones and Silicone-Modified
Materials; ACS Symp. Ser. 2000, 729, 284.
Conclusions
A synthetic route was employed to prepare cationi-
cally polymerizable methyl, cyclopentyl, and cyclohexyl
substituted polysiloxanes. The hydrosilylation of inter-

Macromolecules, Vol. 37, No. 25, 2004
Silane Monomers and Polysiloxanes 9417
(8) Laine, R.; Brick, C.; Kim, S.; Tamaki, R.; Chen, H.; Choi, J.
Abstracts of Papers, 225th ACS National Meeting, 2003, New
Orleans, LA, March 23-27.
(21) Harris, R. K.; Kennedy, J. D.; McFarlane, W. Group IV-
silicon, germanium, tin and lead. In NMR and the Periodic
Table; Harris, R. K., Mann, B. E., Eds.; Academic: New York,
1978; Chapter 10.
(9) Arkles, B. CHEMTECH 1991, 13, 542.
(10) Soucek, M. D.; Wu, S.; Chakrapani, S. Silicones and Silicone-
Modified Materials; ACS Symp. Ser. 2000, 729, 516.
(11) Dworak, D. P.; Soucek, M. D. Prog. Org. Coat. 2003, 47, 448.
(12) Eaborn, C. Organosilicon Compounds; Buttersworth: Lon-
don, 1960.
(22) Doornkamp, T.; Tan, Y. Y. Polym. Commun. 1990, 31, 362.
(23) Chen, J.; Soucek, M. D.; Simonsick, W. J.; Celikay, R. W.
Polymer 2002, 43, 5379.
(24) Nelson, E. W.; Carter, T. P.; Scranton, A. B. J. Polym. Sci.,
Polym. Chem. Ed. 1995, 33, 247.
(13) Smith, R. C.; Angelotti, N. C.; Hanson, C. L. Chemical
Analysis. In Analysis of Silicones; Smith, A. L., Ed.; Wiley:
New York, 1974.
(25) Kennedy, J. P.; Marechal, E. Carbocationic Polymerization;
Wiley: New York, 1982.
(26) Shin, H.; Choi, S.; Kim, J.; Park, I.; Lee, S.; No, J.; Kang, D.
(14) Harris, K. S.; Wang, C. M.; Bowman, C. N. Macromolecules
1994, 27, 650.
Memburein 2003, 13, 246.
(15) Guymon, C. A.; Bowman, C. N. Macromolecules 1997, 30,
1594.
(27) Patra, M.; Sinha, B. K. J. Macromol. Sci., Pure Appl. Chem.
1999, A36, 1459.
(16) ASTM D 1652-97, Standard Test Methods for Epoxy Content
of Epoxy Resins, June 1997.
(28) Doornkamp, A. T.; Tan, Y. Y. Polym. Commun. 1990, 23,
2474.
(17) Smith, A. L.; Angelotti, N. C. Spectrochim. Acta 1959, 15,
412.
(29) Li, L.; Xioa, Z.; Zhang, Z.; Tan, S. Chem. Eng. J. (Amsterdam,
Netherlands) 2004, 97, 83.
(18) Williams, D. E. Nuclear Magnetic Resonance Spectroscopy.
In Analysis of Silicones; Smith, A. L., Ed.; Wiley: New York,
1974; Chapter 11.
(30) Kim, J.; Honma, I. Electrochim. Acta 2003, 48, 3633.
(31) Nordisk, N. A/S, Den. Silicone coating system providing low
friction. Patent number WO9944754, 1999.
(19) Litzow, M.; Spalding, T. Mass Spectrometry of Inorganic and
Organometallic Compounds. In Physical Inorganic Chemistry;
Lappert, M., Ed.; Elsevier Scientific: Amsterdam, 1973;
Monograph 2, Chapter 7
(20) Williams, E. A.; Cargioli, D. Silicon-29 NMR Spectroscopy.
In Annual Reports on NMR Spectroscopy; Webb, G. A., Ed.;
Academic: London, 1979; Vol. 9.
(32) Imperial Chemical Industries Ltd., UK. Organosilicon com-
pounds. Patent number DE2637383, 1977.
(33) Chujo, S.; Rikukawa, M.; Sanui, K.; Ogata, N.; Koizumi, S.;
Hashimoto, T. Polym. J. (Tokyo) 2001, 33, 1.
MA048998W