Synthesis of silsesquioxane urethane hybrid materials by a modified sol-gel process

Article abstract of DOI:10.1039/c5nj00960j

Methacrylate silsesquioxane terminated by urethane acrylate (MASSQ-UA) was prepared via a combination of a sol-gel process and urethane chemistry. The organic and inorganic segments were hybridized at the molecular level. The methacrylate silsesquioxane (MASSQ) was first prepared by hydrolysis and condensation reactions. The molar ratio of water to silane was specifically quantified to control the molecular structure, molecular weight and the distribution of MASSQ. MASSQ-UA was then synthesized by terminating the residual silanol groups in the incomplete MASSQ with pre-prepared mono-adducts of isophorone diisocyanate and 2-hydroxylethyl acrylate. In the second step of the reaction, the residual, but still active, silanol groups were effectively reduced and sterically hindered, overcoming the instability problems that frequently occur in traditional sol-gel-derived silsesquioxane. In addition to the methacrylate groups, isophorone diisocyanate and 2-hydroxylethyl acrylate were hybridized with MASSQ. Unlike traditional heterogeneous filler systems, these hybrids are molecular systems and are homogeneous, optically clear fluids in which the inorganic and organic components are linked by covalent bonds. The molecular structures, molecular weight and polydispersity of the final products were investigated by various methods, including matrix-assisted laser desorption/ionization time-of-flight mass spectrometry and electrospray ionization time-of-flight mass spectrometry. The thermal stability of MASSQ-UA and the water resistance of UV-cured MASSQ-UA-based coatings were improved compared with the MASSQ and UV-cured MASSQ-based coatings. The process developed here can be monitored and controlled and could therefore be used industrially for commercial large-scale production.

Full text of DOI:10.1039/c5nj00960j

NJC
View Article Online
View Journal | View Issue
PAPER
Synthesis of silsesquioxane urethane hybrid
materials by a modified sol–gel process†
Cite this: New J. Chem., 2015,
39, 9789
Sizhe Wang,a Guangli Li,a John J. Chiao,b Z. Jeffrey Wang*b and
Yanwen Yvonne Duan*a
Methacrylate silsesquioxane terminated by urethane acrylate (MASSQ-UA) was prepared via a combination of
a sol–gel process and urethane chemistry. The organic and inorganic segments were hybridized at the
molecular level. The methacrylate silsesquioxane (MASSQ) was first prepared by hydrolysis and condensation
reactions. The molar ratio of water to silane was specifically quantified to control the molecular structure,
molecular weight and the distribution of MASSQ. MASSQ-UA was then synthesized by terminating the
residual silanol groups in the incomplete MASSQ with pre-prepared mono-adducts of isophorone
diisocyanate and 2-hydroxylethyl acrylate. In the second step of the reaction, the residual, but still active,
silanol groups were effectively reduced and sterically hindered, overcoming the instability problems that
frequently occur in traditional sol–gel-derived silsesquioxane. In addition to the methacrylate groups,
isophorone diisocyanate and 2-hydroxylethyl acrylate were hybridized with MASSQ. Unlike traditional
heterogeneous filler systems, these hybrids are molecular systems and are homogeneous, optically clear
fluids in which the inorganic and organic components are linked by covalent bonds. The molecular
structures, molecular weight and polydispersity of the final products were investigated by various methods,
including matrix-assisted laser desorption/ionization time-of-flight mass spectrometry and electrospray
ionization time-of-flight mass spectrometry. The thermal stability of MASSQ-UA and the water resistance
of UV-cured MASSQ-UA-based coatings were improved compared with the MASSQ and UV-cured
MASSQ-based coatings. The process developed here can be monitored and controlled and could
therefore be used industrially for commercial large-scale production.
Received (in Montpellier, France)
17th April 2015,
Accepted 28th August 2015
DOI: 10.1039/c5nj00960j
www.rsc.org/njc
Hybrid materials can be categorized into two classes based
on the nature of the hybrid interface.16 Traditional matrix–filler
Introduction
Organic–inorganic hybrid materials have attracted much interest systems, particularly polymeric matrix–nanofiller systems (nano-
over the last three decades.1,2 This interest in organic–inorganic composites), are class I hybrid materials that use organic polymers
hybrid materials results from the possibility of using the advantages as the matrix and inorganic nanoparticles, nanofibres or nanoplates
of both the organic and inorganic materials and discovering as fillers. These fillers, such as clay,17,18 silica19,20 and other inorganic
new properties that benefit from the synergy of these components.3 species, can remarkably reinforce the physical and chemical
The controls on the interfacial interactions between the organic and performance of polymers even at low concentrations. The
inorganic components are thus a key issue. These organic–inorganic huge specific surface area and resulting interactions between
hybrid materials provide many possibilities for the design of materi- the organic and inorganic phases frequently generate new
als with tailor-made physical and chemical properties and processing properties. However, these nanocomposites are still hetero-
requirements.4 A number of highly functional, high-performance geneous systems. Incompatibility remains an issue if the surface
organic–inorganic hybrid materials have been developed and their modification of the nanofillers is not carried out properly.
potential applications have been exploited in various fields, such as Consequently, self-aggregation and/or agglomeration of the
renewable energy,5,6 catalysis,7 electronics,8,9 optics,10,11 biomedical nanoparticles often occur, offsetting the advantages of the
materials12,13 and as functional coatings.14,15
nanocomposites. In addition to issues of material instability,
many desirable physical or chemical properties may not be
achieved in nanoparticle-filled composites. For example, the
large aggregated and/or agglomerated particles can cause severe
light scattering in optical applications and therefore low optical
transparency, making the materials potentially unsuitable for
most optical applications.21
a College of Chemistry and Molecular Sciences, Wuhan University, Wuhan, 430072,
P. R. China. E-mail: yduan@whu.edu.cn; Fax: +86 027 68753383
b NEO-SITECH Enterprises Limited, 495 Plainfield Ave., Berkeley Heights, NJ 07922,
USA. E-mail: jeffreywang.neo0731@yahoo.com
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5nj00960j
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem., 2015, 39, 9789--9799 | 9789

View Article Online
Paper
NJC
Sol–gel chemistry provides an important method of preparing of polyhedral oligomeric silsesquioxane is time-consuming and
class II hybrid materials.3 The greatest advantage of sol–gel produces a low yield; the cost in terms of both energy and
chemistry is that hybrid materials can be designed and prepared human resources is high.31,32 Thus its commercial applications
at the molecular level,22 avoiding phase separation, light scattering are very restricted.
and many other issues that often occur in the heterogeneous
class I systems.
It would therefore be desirable to develop an industrially
viable method of preparing reproducible organic–inorganic
However, the industrial application of the traditional sol–gel hybrid materials with a stable hybrid interface and the thermal
processes used to produce hybrid materials is limited by two and hydrolytic stability required by customers. In this study,
intrinsic issues. First, the hydrolysis and condensation reactions concentrated efforts were made to solve these problems by
are affected by many different factors, including the stoichiometric taking the two approaches illustrated in Scheme 1. In step 1,
ratio of the reactants, the reaction medium, the type of catalyst, the the stoichiometric amount of water was controlled to optimize
reaction temperature, the type of alkyl group in the precursor and the degree of hydrolysis and the condensation reactions,
other processing conditions.23–25 These frequently cause poor thereby controlling the molecular weight and structures of
product reproducibility in terms of molecular weight and molecular the target silsesquioxane molecules and, consequently, the
structure. Second, it is difficult to drive both the hydrolysis and viscosity of the resulting products. In step 2, the silanol groups
condensation reactions to completion. Consequently, the unhydro- remaining in the incompletely condensed silsesquioxane mole-
lysed alkyoxide groups or uncondensed hydroxyl groups in the final cules were terminated using isocyanate-functional chemicals as
products become the cause of poor thermal and hydrolytic stability the terminating agents. Urethane chemistry has been widely
in the materials produced. Many researchers have explored adopted in industrial applications and is thus a reliable and
sol–gel-derived hybrid materials. Yoldas26 studied the kinetics controllable tool. Step 2 may achieve three goals: (1) making
of metal–organic derived systems. He concluded that the hydro- the reactions controllable and shortening the reaction time;
lysis and condensation reactions occur almost simultaneously (2) linking the organic and inorganic portions via covalent
and are therefore complicated and hard to control. Delattre bonds and then forming organic–inorganic hybrid materials
et al.23 studied the hydrolysis and condensation of methacryloxy- at the molecular level; and (3) opening up an alternative path-
propyltrimethoxysilane under acidic conditions using NMR way to introduce an additional organic functional group into
spectrometry. They found that the hydrolysis and condensa- the target hybrid materials, thus providing more options to
tion reactions were incomplete, even after 14 days. Hench and manipulate the interfacial interactions between the organic
co-workers27,28 reported that the silanol groups did not condense and inorganic components/phases and, ultimately, generating
completely unless the reaction temperature exceeded 400 1C, new types of materials.
whereas the organic constituents in hybrid materials could
hardly survive these temperatures.29
As an example, we synthesized a methacrylate silsesquioxane
terminated by urethane acrylate (MASSQ-UA) to demonstrate
One widely considered solution for better control and stability is that a series of UV-curable, molecular organic–inorganic hybrid
to produce perfectly defined inorganic nano-building blocks, such materials could be prepared in a reliable process that could
as polyhedral oligomeric silsesquioxane.16 As reported previously, be adopted industrially. First, a methacrylate silsesquioxane
silsesquioxane is one of the most promising sol–gel-derived hybrid (MASSQ) was prepared by hydrolysis and the condensation of
materials with a number of characteristic attributes, including the 3-(methylacryloxy)propyltrimethoxysilane (MAPTMS). MASSQ-UA
ability to introduce various organic and inorganic functional was then synthesized by the reaction between the residual
groups on the resulting structures.30,31 However, the preparation silanols in incompletely condensed MASSQ and mono-adducts
Scheme 1 Synthetic route of MASSQ-UA, a multi-functional hybrid material.
9790 | New J. Chem., 2015, 39, 9789--9799
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
of isophorone diisocyanate and 2-hydroxyethyl acrylate (IPDI–HEA). of dilute hydrochloric acid added and the resulting water to silane
Compared with most preparations of silsesesquioxanes,33,34 the total molar ratios. The resulting product 1 was selected, labelled as
time for the preparation of MASSQ-UA was short, o20 h in total. The MASSQ, and used in further preparation procedures.
products were characterized by Fourier-transform infrared (FTIR)
Mono-adducts of isophorone diisocyanate and 2-hydroxyethyl
spectrometry, proton nuclear magnetic resonance (1H-NMR) spectro- acrylate. DBTDL (0.0266 g, 0.10 wt%) and IPDI (17.28 g, 0.0777 mol)
metry, matrix-assisted laser desorption/ionization time-of-flight mass were charged and stirred in a dry reaction flask equipped with a
spectrometry (MALDI TOF MS), electrospray ionization time-of-flight mechanical stirrer, a reflux condenser and a pressure-equalizing
mass spectrometry (ESI TOF MS) and gel-permeation chromato- addition funnel to form a homogeneous liquid. A pre-mixed
graphy combined with static laser light scattering (GPC-LLS). homogeneous solution of 4-methoxyphenol (0.0909 g, 0.34 wt%)
The thermal stability and water resistance were determined using and HEA (9.026 g, 0.0777 mol) was added dropwise to this stirred
testing methods designed in this laboratory.
liquid over about 30 min. The reaction was exothermic and the
reaction temperature was kept at o70 1C by controlling the rate of
addition. The reaction mixture was then stirred at 65–70 1C using
an electric heating mantle. The reaction took about 2 h and was
considered to be complete when the reaction mixture reached the
theoretical NCO value of the IPDI–HEA mono-adducts (12.3%) and
then remained the same at further reaction times. The free NCO
value was measured by back-titration with dibutylamine (ASTM
D2572). A colourless fluid (IPDI–HEA) was obtained at a yield
of 97.9%.
Methacrylate silsesquioxane terminated by urethane acrylate.
DBTDL (0.136 g, 0.10 wt%) and MASSQ (110.0 g, 80.7 wt%)
were added to IPDI–HEA (25.88 g, 19.2wt%) at a controlled rate
with continuous mechanical stirring, maintaining the reaction
temperature at o70 1C. After the addition, the reaction mixture
was stirred for 5 h at 65–70 1C. During this process, the value of
free isocyanate in the reactants was monitored using back-
titration with dibutylamine.
Experimental
Materials
Isophorone diisocyanate (IPDI) was purchased from Alfa Aesar.
2-Hydroxyethyl acrylate (HEA) was purchased from Shanghai
D&B Chemical Lot (Shanghai, China). Dibutyltin dilaurate
(DBTDL) (used as a catalyst) and 4-methoxyphenol (used as an
inhibitor) were purchased from Sigma-Aldrich. 3-Methacryloxy-
propyltrimethoxysilane (MAPTMS) was purchased from WD
Silicone Co. Ltd (Hubei, China). Other chemicals were supplied
by Sinopharm Chemical Reagent Co. Ltd (China). 1-Hydroxy-
cyclohexyl phenyl ketone (Irgacure 184) was acquired from
BASF. 1,6-Hexanediol diacrylate was purchased from Nanjing
Jinlu Chemical Co. Ltd (China). All of the materials were used as
received without further purification.
Preparation procedures
As we did not initially know the amount of uncondensed
silanol groups in MASSQ, a few pre-experiments were carried
out to determine the most appropriate ratio of MASSQ and
IPDI–HEA by using the change in free –NCO groups as a guide.
If the value of free –NCO groups rapidly reached zero during the
reaction, it was concluded that there was an excess amount of
residual silanol in the reactants. Additional IPDI–HEA was then
added and the reaction continued at 65–70 1C for a further
hour. Conversely, when the value of free –NCO remained
40.3% (vs. the total weight of reactants) without any change
for 20 min, it was assumed that excess IPDI–HEA had been
added. In these cases, a suitable amount of HEA was then added and
the reaction was held at 65–70 1C for an additional hour (Fig. 1).
These checks and corrections were repeated until the value of free
isocyanate remained in the range 0–0.3%. At this point, the residual
silanol in the incompletely condensed MASSQ were considered to
have been consumed or, in other words, the incompletely condensed
MASSQ had been effectively capped by IPDA–HEA. The product was
then considered to be sufficiently condensed and end-capped to
have become chemically inert. The resulting organic–inorganic
hybrid material was cooled down to room temperature naturally.
A colourless or pale yellow transparent fluid was obtained at a yield
of 96.0% and was labelled as MASSQ-UA.
Methacrylate silsesquioxane. MASSQ was prepared through
the hydrolysis and condensation reactions of MAPTMS with
hydrochloric acid as the catalyst. Ethanol was used as the
reaction medium. To study the influence of the water to silane
molar ratio at constant pH, different amounts of 0.1 mol LÀ1
hydrochloric acid were added (Table 1). The general procedure
was as follows. MAPTMS (150.0 g, 0.604 mol) was charged and
stirred in a dry reaction flask equipped with a mechanical
stirrer, a reflux condenser and a pressure-equalizing addition
funnel. A homogeneous mixture of ethanol (70 mL) and a
measured amount of dilute hydrochloric acid (0.1 mol LÀ1
)
were added to this stirred liquid through the dropping funnel
over a period of 30–60 min. The reaction mixture was stirred at
60 Æ 5 1C for 4 h and then transferred to a dry round-bottomed
flask to remove the ethanol and other volatile components by
rotary evaporation for 4–6 h at 78 1C. Table 1 shows the amounts
Table 1 Silsesquioxane samples prepared with different water to silane
molar ratios. Different amounts of 0.1 M hydrochloric acid were added to
150 g of silane
Mass of dilute
hydrochloric acid (g)
Water to silane
molar ratio
Product
Measurements
1 (MASSQ)
32.97
164.9
329.7
659.5
989.2
3
15
30
60
90
2
3
4
5
Fourier-transform infrared spectrometry. FTIR spectra were
recorded on a Nicolet iS10 Fourier-transform infrared spectrometer
with KBr plates. The sample was first dissolved in chloroform and
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem., 2015, 39, 9789--9799 | 9791

View Article Online
Paper
NJC
Thermal stability tests. For the thermal stability tests, MASSQ
(20 mL) and MASSQ-UA (20 mL) were poured into two beakers and
labelled MASSQ–T and MASSQ-UA–T, respectively. The samples
were placed in an oven and covered by laboratory-made dark cups,
which had a small vent to maintain contact with the atmosphere
while avoiding exposure to light. The samples were stored at 65 1C
for 7 days and the molecular masses before and after the thermal
stability test were measured by GPC-LLS. This test was repeated on
three batches of the products.
Gel-permeation chromatography-laser light scattering. GPC-LLS
characterization was performed at 35 1C on a Waters 2690D
instrument equipped with Waters Styragel HR1 and HR4 columns,
a 410 differential refractometer detector and a Wyatt DAWN EOS
light-scattering detector. The eluent was THF (HPLC-grade, Fisher)
at a flow-rate of 0.6 mL minÀ1. The refractive index increment
Fig. 1 Percentage of NCO in the reaction mixture during the preparation
of MASSQ-UA. If the amount of free –NCO remained 40.3% without
changing for 20 min, then a suitable amount of HEA was added.
then dispersed on the KBr plate. The FTIR spectra were obtained (dn/dc) was measured on an Optilab DSP (Wyatt Technology)
from the plates after the evaporation of chloroform. instrument. The data were collected and processed using Astra
Proton nuclear magnetic resonance. 1H-NMR spectra were software (Wyatt Technology).
recorded using a Mercury VX-300 spectrometer (Varian, USA)
Thermogravimetric analysis. TGA was carried out on a
using tetramethylsilane as the internal reference and CDCl3 as Setaram Setsys-16 Thermo Gravimetric Analyzer. The heating
the solvent.
programme was set at a rate of 10 1C minÀ1 from room
Matrix-assisted laser desorption/ionization time-of-flight temperature to 800 1C in an air flow atmosphere. Both MASSQ
mass spectrometry. MALDI TOF MS was performed using an and MASSQ-UA were measured.
Applied Biosystems Voyager-DE STR device with delayed ion
Water resistance tests. Water resistance tests were carried
extraction. The instrument was equipped with a pulsed nitrogen out on UV-cured product-based coatings to evaluate the water
laser (l = 337 nm, pulse width = 3 ns, average power 20 Hz = 5 mW). resistance of the products. To prepare the coatings, MASSQ was
The extraction voltage in the TOF analyser was 20 kV. Samples were thoroughly mixed with 1,6-hexanediol diacrylate and Irgacure
measured in the reflector, positive ion mode. Either 2,5-dihydroxy- 184 at a mass ratio of 15 : 10 : 1. The mixture was labelled A.
benzoic acid or a-cyano-4-hydroxycinnamic acid was used as the MASSQ-UA was processed in the same way and labelled B.
matrix material. The samples were dissolved in acetonitrile and Mixtures A and B were spread in triplicate onto the rough
the resulting sample solutions were mixed with the matrix solution surfaces of ground glass plates with an applicator and then
at a ratio of 1 : 1 (v/v). A 1 mL sample of each mixture was applied to immediately UV-cured under a 2 kW UV lamp at a distance of
the sample targets and dried.
15 cm for 2 min to ensure that the mixture was completely
Electrospray ionization mass spectrometry. ESI TOF MS data cured. The adhesion of these cured coating films to the ground
for MASSQ and MASSQ-UA were collected on a Bruker TOF mass glass was pre-assessed according to the Paints and Varnishes-
spectrometer. The reaction intermediates were prepared in a 1 : 1 Cross Cut Test (GB/T 9286-1998). All the cured samples
(v/v) mixture of MeOH and water. In the positive ion mode, the were half-soaked in water at room temperature and allowed to
ions detected in the ESI TOF experiment were typically [M + H]+ stand for 48 h. These tests were repeated on three batches of
ions with little or no fragmentation. ESI MS measurements for products.
IPDI–HEA were performed on a Shimadzu LCMS-2010EV mass
spectrometer (Tokyo, Japan). Data were acquired and processed
using LCsolution 5.42 SP4 software (Tokyo, Japan). The samples
were dissolved in acetonitrile before centrifugation and were
then directly injected into the spectrometer.
Results and discussion
Structural studies
Ultraviolet-visible spectrophotometry. UV-visible spectra were Scheme 1 shows the synthetic route for MASSQ-UA. MASSQ
collected on a TU-1901 UV-visible spectrophotometer (PGeneral, was prepared by the hydrolysis and condensation of MAPTMS,
China) with two quartz cuvettes. Samples 1–5 and MASSQ-UA were R1Si(OCH3)3, where R1
= 3-methacryloxypropyl. Previous
measured without dilution. Pure water was used as a reference research35 has shown that sol–gel derived silsesquioxanes can
material. The absorption spectra were measured at 210–800 nm at be a mixture of species with the general formula Tn(OH)x(OCH3)y,
1 nm intervals.
where T = R1SiO[1.5À(x+y)/2n] (n = 4, 5, 6,... and x, y = 0, 1, 2,...),
Viscosity. Viscosity measurements of samples 1–5 and including a fully condensed polyhedral known as polyhedral
MASSQ-UA were made using a Brookfield LVDV-II + Pro viscometer oligomeric silsesquioxane and incompletely hydrolysed/condensed
equipped with a spindle SC4-25 and a small sample adapter at 50 1C. products. The MASSQ-UA was obtained by the reaction between the
The samples were kept at 50 1C for 1 h before measurement. The residual silanol groups of MASSQ and the isocyanate groups of
viscosity of each sample was measured in triplicate at five different IPDI–HEA. The general formula of MASSQ-UA is denoted as
shear rates from the minimum to the maximum shear rate.
Tn(OH)xÀz(OCH3)yRz2 [R2 = –(OH + IPDI–HEA)].
9792 | New J. Chem., 2015, 39, 9789--9799
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
More significantly, this work was intended to demonstrate a are in slight excess, which is consistent with the results from the
powerful approach to introducing a variety of inorganic and/or titrations. Although residual hydroxyl groups are still present, as
organic functionalities to the target products. For the inorganic indicated by the peak near 3467 cmÀ1, their relative peak area
component, many different metal or semi-metal alkoxide precursors was reduced compared with MASSQ. These experimental results
provide the opportunity to introduce various inorganic atoms, indicate that the hydroxyl groups cannot react further with the
including Si, Ti, Zr, Al and B,36 to the hybrid materials. In this isocyanate, possibly due to steric hindrance effects caused by the
way, desirable properties could be designed for optical, electronic, large IPDI–HEA and/or methacryloxypropyl groups.
mechanical and photoelectric applications. For the organic
Fig. 3a shows the 1H-NMR spectrum of MASSQ. The
component, both the R1 group in the precursor, R1Si(OCH3)3, chemical shifts are 6.10 (CH2Q), 5.55 (CH2Q), 4.11 (–OCH2–),
and the R2 group (the IPDI–HEA adducts in this instance) are 3.82 (Si–O–CH2–), 3.68 (–CH2–O– in ethanol), 3.55 (CH3–O–Si),
variable sites for introducing different functional groups, such 3.42 (CH3–O– in methanol), 1.92 (–CH3), 1.79 (–CH2–), 1.22
as UV-curable acrylate or methacrylate, hydrophobic fluorinated (CH3– in ethanol), 0.72 (–CH2–Si) and 7.34 (CHCl3 in the solvent).
alkyl groups, or hydrophilic ethoxy or phosphate groups. Thus This spectrum shows that there is also a small quantity of residual
particular functionalities or properties could be designed and alkoxy groups in the MASSQ, potentially due to the steric hindrance
produced for customers.
effect, as the hydrolysis reaction in the sol–gel process is often
Fig. 2 shows the FTIR spectra of IPDI–HEA, MASSQ and barely complete.
MASSQ-UA. There is a broad absorption band in the spectrum
Fig. 3b is the 1H-NMR spectrum of IPDI–HEA. The chemical
of IPDI–HEA, with a maximum at 3354 cmÀ1, resulting from the shifts are 6.471–6.413, 5.879–5.845 (CH2Q), 6.190–6.098
amine group in CO–NH. The peak at 2266 cmÀ1 is the absorption (–CHQ), 4.335–4.304 (–CO–CH2–CH2–O–), 3.847–3.663 (–CH
for isocyanate groups. The peak at 1530 cmÀ1 is attributable to on the cyclohexane adjacent to N), 3.521–3.030 (–CH2– adjacent
the vibration of CO–NH.
to –NCO), 1.866–1.505 (–CH2– in the cyclohexane), 1.251–1.030
The presence of a broad absorption band from 3000 to (–CH3 linked to the chiral carbon atom), 0.950 (–CH3 linked to
3800 cmÀ1 with a maximum at 3467 cmÀ1 in the spectrum of the cyclohexane) and 7.281 (CHCl3 in the solvent). The appearance
MASSQ represents the hydroxyl group and indicates an incomplete of NH–CO proton peak demonstrates that the expected reaction
condensation reaction. The peaks at 2956, 2928 and 2894 cmÀ1 between IPDI and HEA has occurred.
are the stretching vibration peaks of CH3 and CH2. The peak
Fig. 3c is the 1H-NMR spectrum of MASSQ-UA. The peaks at
near 1719 cmÀ1 is characteristic of the stretching vibration of 5.8–5.9, 6.1–6.2 and 6.4–6.5 ppm are obvious, which confirms
CQO in the acrylate group. The peak near 1637 cmÀ1 is the presence of acrylate groups in the MASSQ-UA molecular
attributable to the stretching vibration of CQC in the acrylate structure. All these analyses suggest that the –NCO groups have
group and the bending vibration of O–H in water. The peak near
1452 cmÀ1 is from the bending vibration of CH3– and CH2–. The
peak near 1297 cmÀ1 is attributable to the stretching vibration of
C–O and the peaks between 1000–1200 cmÀ1 are caused by the
vibration of Si–O–Si.
In the spectrum of MASSQ-UA, the absorption peak for
isocyanate near 2266 cmÀ1 is very weak, whereas the absorption
peak for the amine groups at 3360 cmÀ1 is strong; this is
significantly different from the spectrum of IPDI–HEA. The
absorption for CO–NH is also seen near 1530 cmÀ1. These
spectral data indicate that the isocyanate groups in IPDI–HEA
have reacted with the hydroxyl groups on MASSQ and that
CO–NH structures have been formed. The isocyanate groups
Fig. 3 1H-NMR (CDCl3) spectra of (a) MASSQ, (b) IPDI–HEA and
Fig. 2 FTIR spectra of (a) IPDI–HEA, (b) MASSQ and (c) MASSQ-UA.
(c) MASSQ-UA.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem., 2015, 39, 9789--9799 | 9793

View Article Online
Paper
NJC
Table 2 Assignment of the main peaks (relative intensities 4 10%) of
MALDI TOF MS for MASSQ
almost reacted completely, as expected, which is consistent
with the results of FTIR analysis.
Soft ionization mass spectrometry, especially MALDI mass
spectrometry37 and ESI mass spectrometry,38,39 has become an
extremely effective tool for the structural elucidation of silses-
quioxanes. The MALDI TOF mass spectra of MASSQ and 802.9019
MASSQ-UA are shown in Fig. 4.
The main peaks with intensities 410% were selected. The
proposed structural formulas corresponding to the main peaks 1302.8185
in Fig. 4 are assigned in Tables 2 and 3.
m/z exp.
Assignment
m/z calc.
Mass error (%)
774.9315
790.8459
T4(OH)4Na+
T4(OH)4K+
775.1911
791.1651
803.2225
945.2311
1115.2710
1133.2816
1303.3215
À0.033
À0.040
À0.040
À0.032
À0.040
À0.039
À0.039
T4(OH)2(OCH3)2Na+
T5(OH)3Na+
T6(OH)2Na+
T6(OH)4Na+
T7(OH)3Na+
944.9262
1114.8239
1132.8382
The data indicate that the MASSQ and MASSQ-UA were
Table 3 Assignment of the main peaks (relative intensities 4 10%) of
MALDI TOF MS for MASSQ-UA
composed of oligomers T4–T7. With the general formula Tn(OH)x-
(OCH3)y for MASSQ and Tn(OH)xÀz(OCH3)y(OR2)z for MASSQ-UA, the
number of loops in the structure can be calculated by eqn (1).
m/z exp.
Assignment
m/z calc.
Mass error (%)
1171.1634
1250.1408
1440.2191
1640.2346
1641.2311
T4(OCH3)3R2K+
T6(OCH3)6K+
1171.3962
1251.3600
1439.4105
1639.6100
1641.5057
À0.020
À0.098
0.056
0.038
À0.017
Loop number = (n + 2 À x À y)/2
(1)
T7(OH)(OCH3)6K+
T5(OH)3R22Na+
T7(OH)2R2Na+
Possible structures could be inferred from the structural formulas.
Fig. 5 shows examples of possible structures for the formula
T4(OH)4 and T7(OH)3 of MASSQ and T4(OCH3)3R2 and T7(OH)2R2
of MASSQ-UA.
ESI TOF MS gives better ionization of silsesquioxane and
more information; however, the resulting spectra (Fig. 6) are
correspondingly more complex. The signals in the ESI spectra may
be species ionized with nH+, nNa+ or nK+, or their combinations
(n = 1, 2,. . .). The composition assignments of MASSQ, IPDI–HEA
and MASSQ-UA with relative intensities 420% were analysed
(Table 4, Table S1 (ESI†) and Table 5, respectively).
The mass spectrum (Fig. 6b) indicates that IPDI–HEA has
been successfully synthesized. However, some by-products were
also found (Table S1, ESI†). As IPDI and IPDI–HEA are both
moisture-sensitive, it was hard to completely prevent the side-
reactions or to purify the product. Encouragingly, the value of
free NCO in the product measured by titration was 12.1%, close
enough to the theoretical NCO value of the IPDI–HEA mono-
adducts (12.3%). According to both the MS and titration results,
it could be inferred that IPDI–HEA was the main component in
the product. As reported previously,40–42 this IPDI–HEA inter-
mediate could be used directly without further processing.
Based on the ESI TOF MS results, although both MASSQ and
MASSQ-UA are still mixtures consisting of compounds with
4–13 silicon atoms, the degree of condensation (or polymeriza-
tion) of the products could be controlled at a certain level by
optimizing the reaction conditions. As a result, these products
have a fine fluidity with a relatively low viscosity.
Among the assignments of MASSQ in Table 4, there are fully
condensed silsesquioxanes such as T8, T10 and T12 structures.21 The
other constituents in the product contain just a few hydroxyl
groups, most with no more than four, as evidenced by the results
of both MALDI TOF MS and ESI TOF MS. Although great efforts
have been made to improve the degree of the hydrolysis and
condensation reactions, it was still hard to drive both reactions
to completion. Thus a hydroxyl eliminating/blocking reaction is
necessary. The results from both mass spectra of MASSQ-UA
indicate that the residual silanol groups were chemically blocked,
while IPDI–HEA was successfully grafted onto the MASSQ.
Optical transparency
Fig. 7 shows a photograph of samples 1–5 and MASSQ-UA.
Samples 1, 2 and MASSQ-UA are visually transparent. Samples
Fig. 4 MALDI TOF mass spectra (reflector positive ion mode) of (a) MASSQ
and (b) MASSQ-UA.
9794 | New J. Chem., 2015, 39, 9789--9799
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
Fig. 6 ESI MS spectra of (a) MASSQ, (b) IPDI–HEA and (c) MASSQ-UA.
Fig. 8 shows the UV-visible spectra of samples 1–5 and
MASSQ-UA with water as the reference. In the range 400–800 nm,
the baseline absorption is almost zero and no peak appears
for either sample 1 (MASSQ) or MASSQ-UA. However, the
absorption of the baseline for samples 2–5 is obviously greater
than zero in the visible light region. These results demonstrate
that a loss of transparency had occurred in samples 3, 4 and 5,
where light scattering had been caused by the suspended
particles in the hazy fluids. Sample 1 (MASSQ) and MASSQ-
UA were also confirmed to be almost colourless and highly
transparent to visible light, suggesting that they are homo-
geneous systems.
Viscosity and homogeneity of MASSQ-UA
Fig. 9 shows the viscosity measurements for samples 1–5
and MASSQ-UA at five different shear rates. The viscosity of
each sample was calculated from the arithmetic mean of the
viscosity at five shear rates. Table 6 gives the viscosity and the
water to silane molar ratio for each sample. The data for
samples 1–5 show that when the water to silane molar ratio
Fig. 5 Typical structures for the assignment of formulas T4(OH)4, T7(OH)3,
T4(OCH3)3R2 and T7(OH)2R2. R1 and R2 are the same as in Scheme 1.
3, 4 and 5 are hazy due to the formation of suspended particles. increased, the viscosity of silsesquioxane significantly increased.
All the samples were initially fluid. After three months of sealed In addition, suspended particles were present in samples 3, 4
storage, samples 1, 2, 3 and MASSQ-UA were still fluid, but and 5. It may be concluded that a higher water to silane molar
samples 4 and 5 had become gels. Thus the water to silane ratio could lead to a larger degree of hydrolysis and condensa-
molar ratio significantly affects the properties of the sol–gel tion of silane. Therefore larger sized molecules were formed and
derived products.
this caused the higher viscosity. This indicates that the molar
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem., 2015, 39, 9789--9799 | 9795

View Article Online
Paper
NJC
Table 4 Composition assignments of the main peaks (relative intensities
4 20%) in the ESI mass spectrum for MASSQ
Table 5 Composition assignments of the main peaks (relative intensities
4 20%) in the ESI mass spectrum for MASSQ-UA
Mass
error (%) m/z exp.
Mass
m/z exp.
Assignment
Ion species
m/z calc.
Assignmenta
Ion species
m/z calc. error (%)
717.1613
731.1092
745.1850
759.1223
775.1620
789.1774
803.1931
817.1806
819.1647
831.2231
833.1941
847.1983
887.1965
903.1715
945.1966
961.1820
959.2113
959.2113
975.2114
973.2275
989.2056
T8
[M + 2H]2+
717.1881 À0.004
731.1998 À0.012
717.3928 T4R2
[M + H + Na]2+ 717.2781
0.016
0.025
0.027
2
T7(OH)(OCH3)6
T8(OH)2
T8(OCH3)2
T4(OH)4
[M + Na + K]2+
[M + H + K]2+
[M + H + K]2+
[M + Na]+
731.4901 T4(OH)3(OCH3)R2
[M + 2H]2+
[M + 2H]2+
[M + Na]+
[M + 2H]2+
731.3055
752.3290
2
745.1713
0.002
752.5328 T4(OCH3)4R2
803.2037 T4(OH)2(OCH3)2
817.2008 T7(OH)(OCH3)R2
831.2366 T7R2
2
759.1870 À0.009
775.1912 À0.004
789.2068 À0.004
803.2225 À0.004
817.2381 À0.007
819.1964 À0.004
831.2538 À0.004
833.2121 À0.002
847.2277 À0.003
887.2241 À0.003
903.2372 À0.007
945.2311 À0.004
961.2050 À0.002
959.2467 À0.004
803.2225 À0.002
817.2733 À0.009
T4(OH)3(OCH3)
[M + Na]+
[M + Na + K]2+ 831.2291
[M + Na + K]2+ 847.2422
0.001
0.035
0.027
T4(OH)2(OCH3)2 [M + Na]+
847.5381 T7(OH)(OCH3)R2
869.5190 T7(OCH3)4R2
887.2107 T9(OH)(OCH3)4
903.1864 T9(OH)2(OCH3)5
973.2444 T10(OH)2(OCH3)4
989.2220 T10(OH)2(OCH3)4
1043.6937 T9(OH)3(OCH3)R2
1075.2567 T12
T4(OH)(OCH3)3
[M + Na]+
[M + 2Na]2+
869.2840
T4(OH)2(OCH3)2 [M + K]+
[M + Na + K]2+ 887.2241 À0.002
T4(OCH3)4
T4(OH)(OCH3)3
T4(OCH3)4
[M + Na]+
[M + K]+
[M + Na + K]2+ 903.2372 À0.006
[M + 2Na]2+
[M + 2K]2+
[M + 2K]2+
[M + 2H]2+
[M + 2Na]2+
[M + 2K]2+
[M + 2Na]2+
[M + 2K]2+
973.2624 À0.002
989.2363 À0.001
[M + K]+
T9(OH)(OCH3)4
[M + Na + K]2+
1043.2797
0.040
T9(OH)2(OCH3)5 [M + Na + K]2+
1075.2785 À0.002
1115.2710 À0.002
1131.2450 À0.001
1143.3023 À0.002
T10(OH)6
T10(OH)6
[M + 2Na]2+
[M + 2K]2+
1115.2483 T12(OH)4
1131.2372 T12(OH)4
1143.2784 T12 (OCH3)4
1161.2740 T12(OH)3(OCH3)3
1245.2929 T13(OH)(OCH3)4
1285.2852 T7(OH)
1303.2807 T7(OH)3
1347.3090 T7(OH)(OCH3)2
1433.3358 T8
T10(OH)4(OCH3)2 [M + 2Na]2+
T10(OH)3(OCH3) [M + 2K]2+
T10(OH)4(OCH3)2 [M + 2K]2+
T10(OH)2(OCH3)4 [M + 2Na]2+
T10(OH)2(OCH3)4 [M + 2K]2+
959.2076
0
1161.2737
0
975.2207 À0.001
973.2624 À0.004
989.2363 À0.003
1003.2520 À0.003
[M + Na + K]2+ 1245.3145 À0.002
[M + Na]+
[M + Na]+
[M + K]+
[M + H]+
[M + Na]+
1285.3109 À0.002
1303.3215 À0.003
1347.3267 À0.001
1433.3689 À0.002
1519.4033 À0.003
1604.4859 À0.007
1003.2206 T10(OCH3)6
1031.1726 T11(OCH3)3
1075.2387 T12
[M + 2K]2+
[M + H + Na]2+ 1031.2783 À0.010
[M + 2H]2+
1075.2785 À0.004
1519.3615 T8(OH)2(OCH3)2
1089.2508 T11(OH)(OCH3)6 [M + Na + K]2+ 1089.2903 À0.004
1604.3786 T13(OH)7(OCH3)2R2 [M + 2K]2+
2
1103.2674 T12(OH)2
1115.2300 T12(OH)4
1129.2444 T12(OH)(OCH3)
1131.2150 T12(OH)4
[M + H + K]2+
[M + 2Na]2+
[M + 2K]2+
1103.2617
0.001
a
R2 represents the group derived from the reaction between –OH and
1115.2710 À0.004
IPDI–HEA shown in Scheme 1.
1129.2475
0
[M + 2K]2+
1131.2450 À0.003
1133.2905 À0.005
1143.3023 À0.004
1143.3023 À0.004
1147.2581 À0.001
1159.2763 À0.003
1159.2763 À0.003
1161.2737 À0.001
1175.2894 À0.001
1177.2868 À0.003
1177.2868 À0.003
1133.2330 T12(OH)(OCH3)3 [M + H + K]2+
1143.2601 T6(OCH3)2
1143.2601 T12(OCH3)4
[M + Na]+
Newtonian rheological behaviour is often observed in a
homogeneous fluid system in which only intermolecular inter-
actions exist. In such cases, the viscosity remains essentially the
same with increasing shear rates. By contrast, a heterogeneous
fluid system often displays a non-Newtonian rheological
behaviour; its viscosity shows shear-thinning phenomena when
both intermolecular and interphase interactions exist in the
system. For samples 1 (MASSQ) and MASSQ-UA, the viscosity
remained almost unchanged with increasing shear rate. These
small and irregular variations in a very narrow range are
believed to be within experimental error. This might be further
evidence indicating that samples 1 and MASSQ-UA are homo-
geneous systems, which is also demonstrated by their high
transparency. However, for samples 4 and 5, the viscosity
decreased significantly when the shear rate increased. This
may suggest that samples 4 and 5 are heterogeneous systems,
as additionally inferred from their hazy appearance and UV-
visible spectra. The viscosity of samples 2 and 3 shows transition
states between samples 1 and 4.
[M + 2Na]2+
1147.2475 T12(OH)5(OCH3) [M + 2K]2+
1159.2384 T6(OCH3)2
1159.2384 T12(OCH3)4
[M + K]+
[M + 2K]2+
1161.2582 T12(OH)3(OCH3)3 [M + 2K]2+
1175.2765 T12(OH)(OCH3)5 [M + 2K]2+
1177.2532 T6(OH)2(OCH3)2 [M + K]+
1177.2532 T12(OH)4(OCH3)4 [M + 2K]2+
1189.2917 T13(OH)3
1191.2681 T13(OH)
[M + H + Na]2+ 1189.3000 À0.001
[M + 2Na]2+
1191.2857 À0.001
1203.2483 T13(OH)(OCH3)2 [M + H + Na]2+ 1203.3157 À0.006
1205.2625 T13(OH)4(OCH3) [M + H + Na]2+ 1205.3131 À0.004
1245.2720 T13(OH)(OCH3)4 [M + Na + K]2+ 1245.3145 À0.003
1263.2743 T7(OH)
1285.2630 T7(OH)
1301.2480 T7(OH)
1303.2661 T7(OH)3
[M + H]+
[M + Na]+
[M + K]+
[M + Na]+
1263.3290 À0.004
1285.3109 À0.004
1301.2849 À0.003
1303.3215 À0.004
1317.2980 À0.001
1319.2954 À0.002
1331.3136 À0.001
1347.3267 À0.003
1433.3689 À0.004
1473.3614 À0.004
1489.3354 À0.002
1491.3720 À0.005
1507.3459 À0.002
1519.4033 À0.004
1535.3772 À0.003
1537.4139 À0.006
1603.4246 À0.005
1621.4194 À0.004
1791.4593 À0.004
1317.2845 T14(OH)3(OCH3) [M + 2K]2+
1319.2649 T7(OH)3
[M + K]+
1331.3010 T14(OH)(OCH3)3 [M + 2K]2+
1347.2841 T7(OH)(OCH3)2
1433.3132 T8
[M + K]+
[M + H]+
[M + Na]+
[M + K]+
[M + Na]+
[M + K]+
1473.3034 T8(OH)2
1489.2984 T8(OH)2
1491.3048 T8(OH)4
1507.3133 T8(OH)4
1519.3359 T8(OH)2(OCH3)2 [M + Na]+
1535.3280 T8(OH)2(OCH3)2 [M + K]+
1537.3269 T8(OH)4(OCH3)2 [M + Na]+
1603.3479 T8(OH)7(OCH3)3 [M + K]+
Thermal stability
The thermal stability was monitored by GPC-LLS. The molecular
weights and polydispersities of MASSQ, MASSQ-UA, MASSQ–T
and MASSQ-UA–T measured by GPC-LLS are shown in Table 7.
The GPC traces are shown in Fig. S1 (ESI†). The increasing
stability was calculated by eqn (2):
1621.3525 T9(OH)
1791.3915 T10
[M + H]+
[M + H]+
ratio of water to silane is crucial to the molecular structure, the
molecular weight and the distribution of MASSQ.
Increasing extent = [Mw(X–T) À Mw(X)]/Mw(X) Â 100%
(2)
9796 | New J. Chem., 2015, 39, 9789--9799
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
Fig. 7 Photograph of samples: 1 (n(water)/n(silane) = 3), 2 (n(water)/
n(silane) = 15), 3 (n(water)/n(silane) = 30), 4 (n(water)/n(silane) = 60), 5
(n(water)/n(silane) = 90) and MASSQ-UA, which was prepared from 1.
Fig. 9 Viscosity of (a) sample 1 (MASSQ), (b) sample 2, (c) sample 3,
(d) sample 4, (e) sample 5 and (f) MASSQ-UA measured at five different
shear rates, changed gradually from the minimum to the maximum shear
rate. The measurements were repeated three times for each sample.
Table 6 Viscosity, appearance and water to silane molar ratio of silses-
quioxane samples 1, 2, 3, 4, 5 and the final product MASSQ-UA
Water to silane
molar ratio
Sample no.
Viscosity (cP)
Appearance
1 (MASSQ)
2
3
4
3
5.1 Â 102
1.1 Â 103
1.6 Â 103
9.1 Â 103
1.9 Â 104
3.7 Â 103
Transparent
Transparent
Slightly hazy
Hazy
Hazy
Transparent
15
30
60
90
3
5
Fig. 8 UV-visible spectra of samples 1–5 and MASSQ-UA. All samples
were measured without dilution. The reference was distilled water. The
region between 400–800 nm was enlarged (inset).
MASSQ-UA
The TGA results also indicated that hybridizing IPDI–HEA
into MASSQ did not lower the heat resistance of the material,
but the maximum thermal degradation temperature of MASSQ-
UA reached 435 1C, 20 1C higher than that of MASSQ (Fig. S2,
ESI†).
where X represents MASSQ or MASSQ-UA, and X–T represents
MASSQ–T or MASSQ-UA–T. Mw is the weight-average molecular
weight of the products.
The values of Mw for MASSQ and MASSQ-UA are about
2.4 Â 103 and 2.7 Â 103 g molÀ1. The Mw increment from MASSQ
Water resistance
to MASSQ-UA suggests that IPDI–HEA has been successfully grafted Transparent coatings were obtained after UV curing of the
on the MASSQ.
MASSQ-based blend A and MASSQ-UA-based blend B on the
After storing both the MASSQ and MASSQ-UA samples at rough surfaces of the ground glass plates. The adhesion tests
65 1C for 7 days, the Mw increased by 130% for MASSQ and 81% showed that none of the coating films was detached, indicating
for MASSQ-UA. This suggests that MASSQ-UA is more thermally good adhesion between coating films and the ground glass.
stable than MASSQ. The probable reason for this is that the Fig. 10a and b are photos of coatings A and B that had been
residual silanols in MASSQ are still active and can continue to half-soaked in water for 48 h. Neither transparent coating
condense to form larger molecules or even gels. Although the changed on exposure to air during the tests (upper half of the
1
FTIR and H-NMR results indicate that there are still residual coatings in Fig. 10a and b). Many scattered white spots
hydroxyl groups and a few alkoxy groups in MASSQ-UA, based appeared on the surface of coating A that was in contact with
on our investigations the hydroxyl groups cannot react with the water (Fig. 10a and c), although no visible change was
excess isocyanate. This means that these hydroxyl and alkoxy observed on coating B (Fig. 10b and d). These results suggest
groups in MASSQ-UA are relatively less chemically reactive due that the water resistance of the MASSQ-UA-based coating B
to steric hindrance.
had been significantly improved compared with that of the
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem., 2015, 39, 9789--9799 | 9797

View Article Online
Paper
NJC
Table 7 GPC-LLS molecular weight information of MASSQ and MASSQ-UA samples before and after thermal stability testing
Mn (g molÀ1
)
Mw (g molÀ1 b
)
PDI (Mw/Mn)c
dn/dc (mL gÀ1 d
)
Increasing extent (%)
a
Sample
MASSQ
2.0 Â 103
3.6 Â 103
1.8 Â 103
2.5 Â 103
2.4 Â 103
5.5 Â 103
2.7 Â 103
4.9 Â 103
1.2
1.5
1.5
1.9
0.063
0.069
0.070
0.070
—
130
—
MASSQ–T
MASSQ-UA
MASSQ-UA–T
81
a
b
c
Mn represents number-average molecular weight. Mw represents the weight-average molecular weight. PDI represents the polydispersity.
dn/dc represents the refractive index increment.
d
product easily formed a hazy suspension or, even worse, a gel.
Although it takes days or even weeks to prepare silsesquioxane
conventionally,33,34 the overall time of the reaction process in this
method was relatively short, usually o20 h. This significant advance
in processing time should positively influence the economic viability
of the industrial production of these materials. The rheological
behaviour suggests that MASSQ-UA is hybridized at the molecular
level and is homogeneous, consistent with its high optical transpar-
ency. It was also confirmed that the thermal stability of MASSQ-UA
and the water resistance of the UV-cured MASSQ-UA-based formula-
tion were significantly improved compared with MASSQ prepared by
the conventional sol–gel process.
In addition, and more interestingly, the methodology
developed here may provide a useful tool for the industrial
design and production of organic–inorganic hybrid materials at
the molecular level. The new methodology also may provide a
broad pathway to graft a variety of desirable organic functional
groups to their hybrid materials, thus providing an effective
method to manipulate the interfacial interactions between
organic and inorganic components or phases.
Further work will be focused on preparing target molecules
Fig. 10 Appearance of (a) coating A (MASSQ-based coating) and (b) coating B
(MASSQ-UA based coating) after soaking in water for 2 days. Only the
lower halves of the two coatings were soaked in water. The enlarged view
(c. 4Â magnification) of the region in white squares on (a) and (b) is
displayed in (c) and (d), respectively.
with more structural specificity and a narrower polydispersity
by various controls and the use of this hybrid material in
optical, optoelectronic and microelectronics applications.
Acknowledgements
MASSQ-based coating A. This is also attributed to the better
chemical stability of MASSQ-UA.
The authors thank Dr Chris W. Miller for generous advice and
help. Financial support from NEO SiTech Enterprises Ltd is
gratefully acknowledged.
Conclusion
Notes and references
An organic–inorganic hybrid material has been designed at the
molecular level. The organic and inorganic portions were
linked via two types of covalent bond: Si–C and Si–O–C. A
modified sol–gel process was developed to prepare this material.
A UV-curable methacrylate silsesquioxane terminated by urethane
acrylate (MASSQ-UA) was synthesized using the mono-adducts
IPDI and HEA as the reacting agents to terminate the incom-
pletely condensed MASSQ. MASSQ was pre-prepared from 3-metha-
cryloxypropyltrimethoxysilane via specific hydrolysis and
condensation reactions by strictly controlling the stoichiometry of
silane and water. It was demonstrated that the stoichiometry used in
the preparation of the silsesquioxane was crucial in controlling the
molecular structures and molecular weights of the silsesquioxanes
and, ultimately, of the final products. Without this control, the
1 C. Sanchez, P. Belleville, M. Popall and L. Nicole, Chem. Soc.
Rev., 2011, 40, 696–753.
2 Y. Chujo, Curr. Opin. Solid State Mater. Sci., 1996, 1,
806–811.
3 U. Schubert, N. Huesing and A. Lorenz, Chem. Mater., 1995,
7, 2010–2027.
´
4 C. Sanchez and P. Gomez-Romero, Functional hybrid materials,
Wiley-VCH, Weinheim, 2004.
5 N. Karousis, J. Ortiz, K. Ohkubo, T. Hasobe, S. Fukuzumi,
´
A. Sastre-Santos and N. Tagmatarchis, J. Phys. Chem. C,
2012, 116, 20564–20573.
6 C. Qu, H. Zhang, F. Zhang and B. Liu, J. Mater. Chem., 2012,
22, 8203–8207.
9798 | New J. Chem., 2015, 39, 9789--9799
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
´
7 U. Dıaz, D. Brunel and A. Corma, Chem. Soc. Rev., 2013, 42, 25 C. J. Brinker and G. W. Scherer, Sol–gel science: the physics
4083–4097.
and chemistry of sol–gel processing, Academic Press, Waltham,
8 L. Yu, Y. Lee, X. Ling, E. J. Santos, Y. C. Shin, Y. Lin,
1990.
M. Dubey, E. Kaxiras, J. Kong and H. Wang, Nano Lett., 26 B. E. Yoldas, J. Non-Cryst. Solids, 1984, 63, 145–154.
2014, 14, 3055–3063.
9 Y. Li, J. Gong, M. McCune, G. He and Y. Deng, Synth. Met.,
2010, 160, 499–503.
27 L. L. Hench and S. H. Wang, Phase Transitions, 1990, 24–26,
785–834.
28 S. Wang, PhD thesis, University of Florida, US, 1988.
10 B. Lebeau and P. Innocenzi, Chem. Soc. Rev., 2011, 40, 29 L. L. Hench and J. K. West, Chem. Rev., 1990, 90, 33–72.
886–906.
30 R. H. Baney, M. Itoh, A. Sakakibara and T. Suzuki, Chem.
Rev., 1995, 95, 1409–1430.
31 D. B. Cordes, P. D. Lickiss and F. Rataboul, Chem. Rev.,
2010, 110, 2081–2173.
11 D. Pletsch, F. Da Silveira Santos, F. S. Rodembusch, V. Stefani
and L. F. Campo, New J. Chem., 2012, 36, 2506–2513.
12 R. Y. Kannan, H. J. Salacinski, P. E. Butler and A. M. Seifalian,
Acc. Chem. Res., 2005, 38, 879–884.
32 X. Cheng, S. Liu and W. Shi, Prog. Org. Coat., 2009, 65, 1–9.
13 C. Dong, G. He, H. Li, R. Zhao, Y. Han and Y. Deng, 33 H. Dong, M. A. Brook and J. D. Brennan, Chem. Mater., 2005,
J. Membr. Sci., 2012, 387–388, 40–47. 17, 2807–2816.
14 J. Godnjavec, B. Znoj, N. Veronovski and P. Venturini, Prog. 34 X. Zhang, L. Hu, D. Sun and W. Zhao, J. Mol. Struct., 2008,
Org. Coat., 2012, 74, 654–659. 872, 197–204.
15 D. Carboni, A. Pinna, L. Malfatti and P. Innocenzi, New 35 P. Eisenberg, R. Erra-Balsells, Y. Ishikawa, J. C. Lucas,
J. Chem., 2014, 38, 1635–1640.
16 F. Ribot and C. Sanchez, Comments Inorg. Chem., 1999, 20,
327–371.
H. Nonami and R. J. J. Williams, Macromolecules, 2002, 35,
1160–1174.
36 J. Wen and G. L. Wilkes, Chem. Mater., 1996, 8, 1667–1681.
37 W. E. Wallace, C. M. Guttman and J. M. Antonucci, J. Am.
Soc. Mass Spectrom., 1999, 10, 224–230.
38 D. R. Bujalski, H. Chen, R. E. Tecklenburg, E. S. Moyer,
G. A. Zank and K. Su, Macromolecules, 2003, 36, 180–197.
39 D. P. Fasce, R. J. J. Williams, R. Erra-Balsells, Y. Ishikawa
and H. Nonami, Macromolecules, 2001, 34, 3534–3539.
40 W. Han, B. Lin, H. Yang and X. Zhang, Polym. Bull., 2012, 68,
1009–1022.
17 J. I. Weon and H. J. Sue, Polymer, 2005, 46, 6325–6334.
18 P. C. LeBaron, Z. Wang and T. J. Pinnavaia, Appl. Clay Sci.,
1999, 15, 11–29.
19 K. A. Mauritz and R. Ju, Chem. Mater., 1994, 6, 2269–2278.
20 Z. H. Huang and K. Y. Qiu, Polymer, 1997, 38, 521–526.
21 Y. Rao and S. Chen, Macromolecules, 2008, 41, 4838–4844.
22 D. A. Loy and K. J. Shea, Chem. Rev., 1995, 95, 1431–1442.
23 L. Delattre, C. Dupuy and F. Babonneau, J. Sol-Gel Sci.
Technol., 1994, 2, 185–188.
41 Q. Gao, H. Li and X. Zeng, J. Coat. Technol. Res., 2011, 8, 61–66.
´
24 R. Ciriminna, A. Fidalgo, V. Pandarus, F. Beland, L. M. Ilharco 42 A. Asif, C. Huang and W. Shi, Colloid Polym. Sci., 2004, 283,
and M. Pagliaro, Chem. Rev., 2013, 113, 6592–6620. 200–208.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem., 2015, 39, 9789--9799 | 9799