Initiating gradient photopolymerization and migration of a novel polymerizable polysiloxane α-hydroxy alkylphenones photoinitiator

Article abstract of DOI:10.1016/j.polymer.2014.06.040

This paper reported a photocleavage type polymerizable organosilicon macromolecular photoinitiator (HHMP-Si-CC), which was synthesized based on traditional photoinitiator 2-hydroxy-1-[4-(2-hydroxyethoxy) phenyl]-2-methyl propan-1-one (HHMP) and amino polysiloxane. HHMP-Si-CC cannot only spontaneously form a concentration gradient in the photopolymerization system, initiate gradient photopolymerization and effectively mitigate inhibition of oxygen, but also overcome the migration of photolysis fragments of the photoinitiator from UV-curable material. Its structure was confirmed by proton nuclear magnetic resonance (¹H NMR), ¹³C NMR, ²⁹Si NMR and Fourier transform infrared spectroscopy (FTIR) and gel permeation chromatography (GPC). The kinetics of photopolymerization of HHMP-Si-CC was studied by real-time infrared spectroscopy (RTIR). Moreover, it was proved by X-ray photoelectron spectroscopy (XPS) and UV absorption that HHMP-Si-CC had relatively good self-floating ability. The polymer initiated by HHMP-Si-CC presented a gradient change in the degree of polymerization, molecular weight, thermostability and glass transition temperature (Tg). The surface morphology of poly(triethylene glycol diacrylate) (PTPGDA) initiated by HHMP-Si-CC was studied by scanning electron microscopy (SEM). More importantly, the migration of photolysis fragments of the HHMP-Si-CC from UV-curable materials was also investigated by high performance liquid chromatography (HPLC). The results showed that the migration more was effectively mitigated by introducing polymerizable groups into the HHMP-Si-CC. HHMP-Si-CC should have potential applications for preparing more environmentally friendly gradient materials.

Full text of DOI:10.1016/j.polymer.2014.06.040

Polymer 55 (2014) 3656e3665
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Initiating gradient photopolymerization and migration of a novel
polymerizable polysiloxane a-hydroxy alkylphenones photoinitiator
,
b
,
, c
Fang Sun a *, Yanxia Li b, Nan Zhang b, Jun Nie a
a State Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, Beijing 100029, PR China
b College of Science, Beijing University of Chemical Technology, Beijing 100029, PR China
c College of Materials Science and Engineering, Beijing University of Chemical Technology, Beijing 100029, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
This paper reported a photocleavage type polymerizable organosilicon macromolecular photoinitiator
(HHMP-Si-CC), which was synthesized based on traditional photoinitiator 2-hydroxy-1-[4-(2-
hydroxyethoxy) phenyl]-2-methyl propan-1-one (HHMP) and amino polysiloxane. HHMP-Si-CC cannot
only spontaneously form a concentration gradient in the photopolymerization system, initiate gradient
photopolymerization and effectively mitigate inhibition of oxygen, but also overcome the migration of
photolysis fragments of the photoinitiator from UV-curable material. Its structure was confirmed by
proton nuclear magnetic resonance (1H NMR), 13C NMR, 29Si NMR and Fourier transform infrared
spectroscopy (FTIR) and gel permeation chromatography (GPC). The kinetics of photopolymerization of
HHMP-Si-CC was studied by real-time infrared spectroscopy (RTIR). Moreover, it was proved by X-ray
photoelectron spectroscopy (XPS) and UV absorption that HHMP-Si-CC had relatively good self-floating
ability. The polymer initiated by HHMP-Si-CC presented a gradient change in the degree of polymeri-
zation, molecular weight, thermostability and glass transition temperature (Tg). The surface morphology
of poly(triethylene glycol diacrylate) (PTPGDA) initiated by HHMP-Si-CC was studied by scanning elec-
tron microscopy (SEM). More importantly, the migration of photolysis fragments of the HHMP-Si-CC
from UV-curable materials was also investigated by high performance liquid chromatography (HPLC).
The results showed that the migration more was effectively mitigated by introducing polymerizable
groups into the HHMP-Si-CC. HHMP-Si-CC should have potential applications for preparing more envi-
ronmentally friendly gradient materials.
Received 23 January 2014
Received in revised form
30 May 2014
Accepted 14 June 2014
Available online 21 June 2014
Keywords:
Gradient polymerization
Polymerizable organosilicon photoinitiator
Migration
© 2014 Elsevier Ltd. All rights reserved.
1. Introduction
intensity [6], applying gradient temperature field [7] or gradient
solution concentration [8], and utilizing magnetic separation [9],
Over the past few years, polymeric gradient materials (PGM)
[1e3] possessing a gradient in composition, microstructure and
property have attracted much attention for investigating the effects
of variation on multiple parameters [4]. PGM exhibit spatial vari-
ations in structure and/or composition, which have been imposed
to obtain unprecedented properties and functionalities that cannot
be realized in homogeneous materials. The use of PGM is contin-
uously growing in industry as reflected by many applications in not
only aerospace, energy, electronics, medical but also optical and
other industries [5]. With the increased attention from researchers,
a wide range of process technologies including changing light
are now available for preparing PGM. However, these techniques
have limitations because of multiple steps, thereby increasing cost
and pollution to environment and decreasing convenience. There-
fore, the development of a simpler, more environmentally friendly
and convenient method for preparing PGM has been a focus of
much research in recent years.
We have proposed a fast, convenient and environmentally
friendly method to prepare molecular weight gradient polymer by
using a polysiloxane benzophenone photoinitiators with good self-
floating ability, leading to the spontaneous formation of a con-
centration gradient of the photoinitiator in a photopolymerization
system. Our previous work [10] reported the effect of silicone chain
length of the polysiloxane benzophenone photoinitiators on the
self-floating ability of the photoinitiator, and the microstructure
and surface property of the molecular weight gradient polymer,
however, the research on the gradient distribution of the degree of
* Corresponding author. College of Science, State Key Laboratory of Chemical
Resource Engineering, Beijing University of Chemical Technology, PO Box 144,
Beijing 100029, PR China. Tel.: þ86 10 64449336; fax: þ86 10 64437802.
E-mail address: sunfang60@yeah.net (F. Sun).
http://dx.doi.org/10.1016/j.polymer.2014.06.040
0032-3861/© 2014 Elsevier Ltd. All rights reserved.

F. Sun et al. / Polymer 55 (2014) 3656e3665
3657
the photopolymerization initiated by the polysiloxane-based pho-
toinitiator and properties of the obtained gradient polymer, and
their internal connection were not involved. The research is very
important to prepare an expected gradient polymer. It is notable
mechanical stirrer and a condenser. The solution was stirred at
ambient temperature (25 ꢁC) for 2 h and then washed three times
with deionized water. The organic layer was dried over Na2SO4, and
filtered, then distilled under vacuum. The crude product was pu-
rified by silica gel (200e300 mesh) column chromatography using
ethyl acetate and methylene chloride (1:20v/v) as an elution and
the yield was 62.5%. The synthesis process is shown in Scheme 1.
FTIR (KBr, cmꢀ1): 3456 cmꢀ1 (eOH), 3062 cmꢀ1 (¼CeH),
that the reported polysiloxane-based photoinitiators are
a
hydrogen-abstraction type photoinitiator system (Norish type II)
that has drawbacks of yellowing, toxicity and mutagenicity derived
from coinitiator amines [11e14]. Therefore, the photocleavage type
(Norish type І) polysiloxane-based photoinitiators that produce
two radicals through an unimolecular cleavage to initiate poly-
merization have been expected as good alternatives. Additionally,
when photoinitiators initiate polymerization, their photolysis
fragments and unreacted molecules might diffuse from UV-cured
materials, resulting in an impact on human health and environ-
ment. One important method to solve the problem is the intro-
ducing polymerizable groups into photoinitiator molecules
[15e17], which can be polymerized into the polymer backbone.
In this work, we reported a novel polymerizable photocleavage
polysiloxane macromolecular photoinitiator (HHMP-Si-CC) that
overcame some drawbacks of the hydrogen-abstraction type
polysiloxane-based photoinitiator and the migration of photolysis
fragments of photoinitiator from UV-cured materials. The self-
floating ability of the photoinitiator was evaluated by the UV ab-
sorption spectra and X-ray photoelectron spectroscopy (XPS). The
migration of photolysis fragments of the photoinitiator from UV-
curable materials was also investigated by high performance
liquid chromatography (HPLC). More importantly, the gradient
distribution of the degree of the photopolymerization initiated by
the photoinitiator and properties of the prepared gradient polymer,
and their internal connection were fully discussed.
2984 cmꢀ1
,
2865 cmꢀ1 (eCH3, eCH2), 1663 cmꢀ1 (>C¼O),
1598 cmꢀ1, 1445 cmꢀ1 (eC6H5), 1359 cmꢀ1 (eSO2e), 1250 cmꢀ1
1020 cmꢀ1 (AreOeC).
，
1H NMR (400 MHz, CDCl3, ppm): 8.03, 7.82, 7.35, 6.84 (AreH),
4.42-4.40 (CH2CH2), 2.45 (ArCH3), 1.62 (CH3CCH3).
2.2.2. Synthesis of NH2-HHMP
HHMP-S (13.7 g, 0.04 mol), A-Si (9.0 g, 0.02 mol) and K2CO3
(19.9 g, 0.14 mol) were dissolved in 300 mL of DMF in a 500 mL
four-necked round bottom flask, which was equipped with a me-
chanical stirrer, a thermometer and a condenser. The solution was
stirred at 110 ꢁC for 24 h. The product was washed three times with
10% NaOH aqueous solution and deionized water, respectively. The
solvent was removed by vacuum distillation. Then the crude
product was purified by silica gel (200e300 mesh) column chro-
matography using ethyl acetate and hexane (1:4v/v) as an elution.
The yield was 58.7%. The molecular weight measured by GPC was
about 897. The synthesis process is shown in Scheme 2.
FTIR (KBr, cmꢀ1): 3342 cmꢀ1 (eOH, eNH), 3068 cmꢀ1 (¼CeH),
2959 cmꢀ1 2875 cmꢀ1 (eCH3, eCH2), 1661 cmꢀ1 (>C]O),
,
1594 cmꢀ1 (eC6H5), 1374 cmꢀ1 (eCH3), 1033 cmꢀ1 (SieOeSi),
1253 cmꢀ1 and 798 cmꢀ1 (SieCH3).
1H NMR (400 MHz, CDCl3, ppm): 8.08, 6.96 (AreH), 3.05e4.17
(CH2CH2), 1.64 (CH3CCH3), 0.09 (SiCH3).
2. Experimental details
2.1. Materials
2.2.3. Synthesis of HHMP-Si-CC
NH2-HHMP (0.62 g, 0.68 mmol), 3-bromopropene (0.33 g,
2.72 mmol) and K2CO3 (1.88 g,13.60 mmol) were dissolved in 20 mL
of acetone in a 100 mL three-necked round bottom flask, which was
Amino polysiloxane (A-Si) (Mn ¼ 450, amino equivalent:
225 g molꢀ1) was from Shin-Etsu Chemical Co. Ltd. (Shanghai,
China).
2-hydroxy-1-[4-(2-hydroxyethoxy)
phenyl]-2-methyl
equipped with
a mechanical stirrer, a thermometer and a
propan-1-one (HHMP), N,N-dimethylaniline, triethanolamine
(TEOA), p-toluenesulfonyl chloride (TsCl), N, N-dimethylfor
mamide (DMF), potassium carbonate (K2CO3), potassium hydroxide
(KOH), methylene chloride (CH2Cl2), 3-bromopropene and ethyl
acetate were purchased from Sinopharm Group Chemical Reagent
Co. (Beijing, China). Tripropylene glycol diacrylate (TPGDA) and
methyl methacrylate (MMA) monomers were purchased from
Eternal Chemical Co. Ltd. (Tianjin, China). All reagents were used as
received without further purification.
condenser. The solution was refluxed at 70 ꢁC for 12 h and washed
three times with deionized water. The organic layer was dried over
Na2SO4, and filtered, then distilled under vacuum. The crude
product was purified by silica gel (200e300 mesh) column chro-
matography using ethyl acetate and hexane (1:4v/v) as an elution
and the yield was 47.3%. The molecular weight measured by GPC
was about 1073. The synthesis process is shown in Scheme 3.
FTIR (KBr, cmꢀ1): 3456 cmꢀ1(eOH), 3079 cmꢀ1 (¼CeH),
2952 cmꢀ1 2872 cmꢀ1 (eCH3, eCH2), 1666 cmꢀ1 (>C]O),
,
1638 cmꢀ1 (eC]Ce), 1601 cmꢀ1 (eC6H5), 1373 cmꢀ1 (eCH3),
1033 cmꢀ1 (SieOeSi), 1259 cmꢀ1and 799 cmꢀ1 (SieCH3).
1H NMR (400 MHz, CDCl3, ppm): 8.08, 6.95 (AreH), 6.03e6.10,
5.32e5.47 (CH2]CH), 3.90e4.21 (CH2CH2), 1.64 (CH3CCH3), 0.90
(SiCH3).
2.2. Synthesis of polysiloxane photoinitiator HHMP-Si-CC
2.2.1. Synthesis of HHMP-S
HHMP (26.9 g, 0.12 mol), TsCl (19.0 g, 0.10 mol) and KOH (22.4 g,
0.40 mol) were dissolved in 300 mL of CH2Cl2 in a 500 mL three-
13C NMR (400 MHz, CDCl3, ppm)
125.3, 114.7, 112.9, 75.2, 68.5, 66.5, 59.5, 44.5, 27.4, 14.2, 0.1.
d: 201.5, 161.5, 145.3, 131.4,
necked round bottom flask, which was equipped with
a
Scheme 1. Synthesis route of HHMP-S.

3658
F. Sun et al. / Polymer 55 (2014) 3656e3665
Scheme 2. Synthesis route of NH2-HHMP.
29Si NMR (400 MHz, CDCl3, ppm)
d
: ꢀ22.2 (Si (CH3)2).
pre-cleaned glass slide and allowed to spread to a diameter of
20 mm to obtain liquid resin film of desired thickness. After
standing for 60 min, the deposited liquid film was then exposed to
high-pressure mercury lamp with 365 nm emission wavelength
(incident light intensity ¼ 50 mW cmꢀ2) for 60 s in aerobic con-
ditions. After exposure, the uncured liquid layer on top was allowed
to swell the bottom gradient film for 20 min to fully develop the
wrinkles to obtain the gradient polymer films, which were
measured by SEM.
2.3. Synthesis of gradient polymer cylinder initiated by HHMP-Si-
CC
HHMP-Si-CC (0.5 wt%, relative to monomers) was dissolved in
TPGDA or MMA monomer, then as-prepared solution was added
into a vertical glass tube with an inner diameter of 6 mm to form
4 cm fluid column. After standing for 60 min, the fluid column was
irradiated with the high-pressure mercury lamp with 365 nm
emission wavelength (incident light intensity ¼ 50 mW cmꢀ2
,
2.5. Characterization methods
recorded by UV Light Radiometer (Photoelectric Instrument Fac-
tory, Beijing Normal University, Beijing, China)). The photoinitiator
content, degree of polymerization, molecular weight, glass transi-
tion temperature (Tg), and thermostability of different vertical level
of the gradient polymer were measured by XPS, Raman spectra,
GPC, DSC and TGA.
2.5.1. Characterization of the photoinitiator
The Fourier transform infrared (FTIR) spectra were recorded
according to a Nicolet 50XC spectrometer (Nicolet, USA) and
scanned between 400 and 4000 cmꢀ1
.
The 1H NMR spectra were recorded on an AV400 unity spec-
trometer (Bruker, Germany) operated at 400 MHz with CDCl3 as a
solvent and tetramethylsilane as an internal standard.
2.4. Synthesis of gradient polymer film in aerobic conditions
Photoinitiators HHMP (4.0 ꢂ 10ꢀ3 mol Lꢀ1) and HHMP-Si-CC
(4.0 ꢂ 10ꢀ3 mol Lꢀ1) were dissolved in the TPGDA respectively. A
measured volume of the UV-resin formulation was dispensed on a
2.5.2. The average molecular weight of the gradient polymer
The average molecular weight of the gradient polymer was
determined by GPC (Water 515-2410, USA). Tetrahydrofuran,
Scheme 3. Synthesis route of HHMP-Si-CC.

F. Sun et al. / Polymer 55 (2014) 3656e3665
3659
1.0 mL minꢀ1, was used as the mobile phase. Linear polystyrenes
(PS) with various molecular weights were used as the calibration
standard.
photoinitiator from UV-curable film. The basic concept for our
approach is schematically illustrated in Scheme 4.
2.7. Photopolymerization
2.5.3. UV absorption and UV degradation of the photoinitiator
The properties of UV absorption and UV degradation were
studied by UV absorption spectroscopy (Hitachi High-Technologies
Corporation, Tokyo, Japan) and scanned between 200 and 500 nm.
TPGDA was employed to investigate kinetics of photo-
polymerizations initiated by HHMP-Si-CC with real-time infrared
spectroscopy (RTIR) (Nicolet 5700, Thermo Electron, USA, equipped
with an extended range KBr beam-splitter and an MCT/A detector)
according to previous reported procedure [10].
2.5.4. The surface morphology and elemental composition of the
polymer
2.8. Raman spectra
The surface morphology of the polymer was observed by SEM
(S-4700 Hitachi) with an accelerating voltage of 20.0 kV.
The elemental compositions of different vertical levels of the
gradient polymer rod (PTPGDA) initiated by HHMP-Si-CC were
investigated by using XPS (Thermo Electron Corporation, Escalab
250, Germany).
Raman spectra of different vertical level of PMMA rod
mentioned in section 2.3 were collected using a Renishaw InVia
reflex Raman microscope to detect degree of photopolymerization.
The Raman scattering was excited with a 785 nm near-infrared
diode laser. Data acquisition covered the spectral range
3000e120 cmꢀ1 with a spectral resolution of 1.5 cmꢀ1 with each
exposure of the CCD detector.
A band whose intensity does not change during polymerization
should be selected as a reference band. In this study, the CeH
stretching band at 2986 cmꢀ1 was selected as a reference band. It is
well known that the C]C stretching mode at 1640 cmꢀ1 can be
influenced during polymerization. Then the intensities for the in-
dividual bands (2986 and 1640 cmꢀ1) were obtained from the peak
areas by the curve fitting software. Care should be taken that the
obtained percentages were based on the assumption that the
linearity of the ratios between uncured and cured states existed.
The percentage of the final double bond conversions (DBC) was
calculated using the following equation:
2.5.5. DSC and thermal analysis
Differential scanning calorimetric (DSC) measurements were
performed in a nitrogen atmosphꢀe1re on 5e15 mg polymer samples
at a heating rate of 20 ꢁC min
using Pyris 1 (Perkin Elmer).
Thermogravimetry analysis (TGA) was performed on STA-449C
thermogravimetric analysis instrument (NETZSCH Instrument Co.,
Germany)ꢀu1nder a nitrogen atmosphere with a heating rate of
20 ꢁC min in the temperature range of 25e500 ꢁC.
2.5.6. HPLC analysis
The HPLC analysis was performed on an Agilent 1100 series
instrument coupled to a variable wavelength detector (VWD).
Samples were separated on a C18 column (5
m
m, 150 ꢂ 4.6 mm i.d.)
with a sample injection volume of 25 L. The mobile phase was
m
DBCð%Þ ¼ ð1 ꢀ Acured=AuncuredÞ ꢂ 100%
where
(1)
mixtures of acetonitrile (A) and water (B). A gradient program was
used as follows: 70% A (0 min), 70% A (5 min), 100% A (15 min). The
mobile phase flow rate was 1.0 mL minꢀ1. Column temperature was
controlled at 25 ꢁC.
Acured ¼ (C]C absorbance area/CeH absorbance area)cured
Auncured ¼ (C]C absorbance area/CeH absorbance area)uncured
.
2.6. Migration of photoinitiator fragment of photoinitiator from UV-
curable film
2.9. Self-floating ability
Photoinitiator HHMP (1.0 ꢂ 10ꢀ3 mol Lꢀ1) and HHMP-Si-CC
(5.0 ꢂ 10ꢀ4 mol Lꢀ1) with the same molarity of the photosensi-
tive group were dissolved in the TPGDA respectively. A measured
volume of the UV-resin formulation was dispensed on a pre-
cleaned glass pane and allowed to spread to a diameter of 20 mm
to obtain a liquid resin film of desired thickness, then covered with
a cover glass to ensure anaerobic conditions. Subsequently, the
deposited liquid film was exposed to the high-pressure mercury
lamp with 365 nm emission wavelength (incident light
intensity ¼ 10 mW cmꢀ2) for 15 min to obtain the cured films. Then
1.0 g of the mashed cured films was extracted with acetonitrile in a
Soxhlet extractor for 96 h. The extraction solution was analyzed by
HPLC to investigate the migration of photolysis fragment of the
The self-floating of the photoinitiator was investigated by UV
absorption spectra. The experimental details were as follows.
HHMP-Si-CC was dissolved in absolute ethyl alcohol to obtain
homogeneous solution with various concentrations 10ꢀ2 mol Lꢀ1
,
10ꢀ3 mol Lꢀ1 and 10ꢀ4 mol Lꢀ1, and the as-prepared solutions were
added into cuvettes to measure the UV absorption. The maximum
absorbance of HHMP-Si-CC was obtained from the UV absorption
spectra. According to the LamberteBeer law, the average absorp-
tion coefficient (ε) of HHMP-Si-CC was calculated to
2.556 ꢂ 104 L molꢀ1 cmꢀ1. Then HHMP-Si-CC solution with con-
centration 10ꢀ3 mol Lꢀ1 was added into a vertical glass tube having
an inner diameter of 3 cm and depth of 25 cm. After standing at
UV
polymer film
acetonitrile
curing
extracting
HPLC
Monomer
Photoinitiator
Glass slide
Scheme 4. Schematic representation of the extraction of the cured film initiated by HHMP and HHMP-Si-CC as a photoinitiator in anaerobic conditions.

3660
F. Sun et al. / Polymer 55 (2014) 3656e3665
various time scales ranging from 0 to 60 min, 1.0 mL aliquots were
carefully taken out from top or bottom in the vertical glass tube and
put into a 10 mL volumetric flask respectively, and then diluted
with 9.0 mL of absolute ethyl alcohol before measuring the UV
absorption. With the maximum absorbance of the UV absorption
spectra, the concentrations of HHMP-Si-CC in the top layer and the
bottom layer at different standing time were calculated according
to the LamberteBeer law.
intensity were also similar due to contain the same molar con-
centration of chromophore.
The UV/vis degradation curves of HHMP-Si-CC in absolute
ethanol are shown in Fig. 4. The results revealed that the intensity
of the UV/vis absorption tended to decrease against the increase of
the UV irradiation time because of the decomposition of HHMP-Si-
CC. Additionally, the pep* transition of HHMP-Si-CC underwent a
blue-shift after irradiation. It may be due to dealkylation process of
HHMP chromophore under UV light irradiation [18].
For comparing, similarly, the concentrations of HHMP in the top
layer and the bottom layer at different standing time were calcu-
lated according to the above procedure.
3.3. Kinetic study of photopolymerization initiated by HHMP-Si-CC
The plots of the double bond conversion vs. time for TPGDA
initiated by HHMP-Si-CC and HHMP with the same molar con-
centration of chromophore are shown in Fig. 5. It was found that
HHMP and HHMP-Si-CC had similar photointiating activities.
Fig. 6 shows that the polymerization rate and the final double
bond conversion increased with the increase of HHMP-Si-CC con-
centration as more free radicals were produced with the higher
concentration of HHMP-Si-CC.
3. Results and discussion
3.1. Synthesis and structural characterization of HHMP-Si-CC
With comparison of with the curve of HHMP, the typical ab-
sorption peaks of C]C were observed at about 3079 and
1638 cmꢀ1. Characteristic absorption peaks of OHe, >C]O, eC6H5,
SieOeSi and SieCH3 appear at 3343, 1659, 1594, 1513, 1036 and
1259 cmꢀ1 respectively in the curve of HHMP-Si-CC, indicating that
HHMP was successfully linked with the A-Si (Fig. 1).
When the light intensity was lower than 30 mW cmꢀ2, the in-
crease in the light intensity resulted in an increase in the poly-
merization rate and final double bond conversion due to more free
radicals produced during irradiation as shown in Fig. 7. However, at
a too high light intensity, the polymerization rate and the final
double bond conversion were reduced. The possible explanation for
the result was the fact that the rapid photodegradation of the
photoinitiator leaded to a lower concentration of propagating
radicals due to an increase in the rate of radicaleradical combina-
tion reactions. At the relatively high light intensities, radical com-
bination reactions could dominate thereby preventing effective
polymerization [14].
Fig. 2 shows that the signals of aromatic hydrogen of HHMP at
6.9e8.0 ppm and allyl peaks at 5.7e6.1 ppm were observed in the
spectrum of HHMP-Si-CC, indicating that A-Si had been linked with
HHMP and 3-bromopropene. After a nucleophilic substitution re-
action, the chemical shifts of eCH2CH2e in HHMP (4.2e4.4 ppm)
were shifted upfield (3.9e4.2 ppm) in HHMP-Si-CC due to the effect
of N atom. The integration ratio of SieCH3 to SieCH2 (peak a' in
Fig. 2) in HHMP-Si-CC was 7.0, which was similar to the corre-
sponding ratio (7.1) of SieCH3 to SieCH2 (peak a in Fig. 2) in A-Si,
and the integration ratio of H-d' to H-c' in HHMP-Si-CC was 2.0,
demonstrating that HHMP and 3-bromopropene linked with both
sides of the A-Si. Additionally, integration ratio verified that poly-
merization degree (n) of A-Si was 3e5. Hence, the NMR estimates
could deduce the molecular weight of A-Si was 392e540 and
HHMP-Si-CC was 889e1037, similar to the GPC results.
3.4. Self-floating ability
Fig. 8 shows the HHMP and HHMP-Si-CC concentrations of the
top layer and bottom layer of the system undergoing different
standing time. With the increase of standing time, the HHMP-Si-CC
concentration of the top layer increased, while the concentration of
the bottom layer decreased, and they all reached equilibrium at
about 50 min, with the migration rate of about 13%. However, the
HHMP concentration of the top layer and bottom layer did not vary
significantly with the increase of standing time. This result indi-
cated that the HHMP-Si-CC had a relatively good self-floating
ability and spontaneously formed a gradient distribution of con-
centration in the polymerization system.
3.2. UV/vis absorption degradation properties
The UV absorption spectra of HHMP and HHMP-Si-CC in abso-
lute ethanol are showed in Fig. 3. The maximum absorption peaks
of HHMP and HHMP-Si-CC all were at 275 nm and absorption
The abundances of C, O and Si on the surface of each segment
from top to bottom layer of polymer rod prepared using HHMP-Si-
CC as an initiator were detected (Fig. 9(a)). The content of Si can
directly reflect the content of the HHMP-Si-CC since the monomer
does not contain Si element. Fig. 9(b) shows that content of Si
decreased gradually from the top to the bottom layer. The content
of Si on the top layer, the second layer, the third layer and the
bottom layer was 14.80%, 8.30%, 5.91% and 3.39%, respectively,
which meant that there was a gradient distribution of HHMP-Si-CC
from the surface down to the bottom in the polymerization system.
Moreover, there was the deviance between the results of UVevis
measurements in Fig. 8 and the XPS analysis, because of the
different medium (the ethanol solution and the UV-curable system
of TPGDA) and measuring methods.
3.5. Gradient polymerization of PMMA initiated by HHMP-Si-CC
Both the UV and XPS measurements confirmed that there was
Fig. 1. FTIR spectra for HHMP (a) and HHMP-Si-CC (b).
an uneven radical photoinitiator concentration distribution in the

F. Sun et al. / Polymer 55 (2014) 3656e3665
3661
Fig. 2. 1H NMR spectra of A-Si and HHMP-Si-CC.
photopolymerization system containing HHMP-Si-CC, which will
affect degree of photopolymerization of the system, thereby influ-
encing properties of the polymer. The gradient polymerization of
the PMMA initiated by HHMP and HHMP-Si-CC was investigated by
Raman spectrum (Fig. 10 and Table 1). It was easily found that for
HHMP system, DBC presented no clear trend from top layer to
bottom layer of PMMA rod, whereas for HHMP-Si-CC system, there
was a gradual decline of the DBC from top layer to bottom layer of
PMMA rod because the concentration of the HHMP-Si-CC
decreased from top layer to bottom layer of PMMA rod. It is well
known that DBC of methacrylate is closely related to the concen-
tration of photoinitiator, and higher concentration of photoinitiator
contributes to the improvement of DBC.
3.6. Molecular weight of gradient polymer
The radical photoinitiator concentration also significantly in-
fluences the polymer molecular weight that increases with the
Fig. 3. UVevis absorption spectra of HHMP and HHMP-Si-CC in absolute ethyl alcohol
with
the
same
molar
concentration
of
chromophore
Fig. 4. UVevis degradation curve of HHMP-Si-CC in absolute ethyl alcohol
(concentration ¼ 1 ꢂ 10ꢀ4 mol Lꢀ1).
(concentration ¼ 2 ꢂ 10ꢀ4 mol Lꢀ1).

3662
F. Sun et al. / Polymer 55 (2014) 3656e3665
Fig. 5. Conversion versus time for the polymerization of TPGDA initiated by HHMP and
Fig. 7. Effect of the light intensity on the polymerization of TPGDA with 0.5 wt%
HHMP-Si-CC.
HHMP-Si-CC
with
the
same
molar
concentration
of
chromophore
(concentration ¼ 9 ꢂ 10ꢀ3 mol Lꢀ1).
HHMP-Si-CC gradually increased from top layer to bottom layer,
indicating the gradual improvement of the thermal stability from
top to bottom layer, while T5% and Tmax for PMMA initiated by
HHMP presented an irregular change. This result was attributed to
the progressive increase in the molecular weight for the PMMA rod
initiated by HHMP-Si-CC from top to bottom layer. The increase of
molecular weight of the polymer enhanced thermal stability of the
polymer. It was also observed from Table 2 that the Tg decreased
from the top layer to the bottom layer of PMMA initiated by HHMP-
Si-CC, while the Tg of PMMA initiated by HHMP varied irregularly.
This may be attributed to the decrease of the HHMP-Si-CC con-
centration from the top layer to the bottom layer of PMMA. In
general, high photoinitiator concentration can increase the degree
of branching of the polymer, while branching chains and acrylate
pendant groups of PMMA are rigid groups, which confine the chain
movement, thereby enhancing Tg.
decrease of photoinitiator concentration. The PMMA rod prepared
using HHMP and HHMP-Si-CC respectively, as a photoinitiator, was
cut into four small segments evenly to detect the effect of the
gradient distribution of HHMP-Si-CC concentration on the molec-
ular weight of the polymer. Fig.11 shows that a progressive increase
in the molecular weight for the PMMA rod initiated by HHMP-Si-CC
from top to bottom layer. The result was attributed to the gradient
distribution of concentration of HHMP-Si-CC. However, the mo-
lecular weight for the PMMA rod initiated by HHMP did not present
an obviously gradient variation from top layer to bottom layer.
3.7. Thermostability and Tg of gradient polymer
Above research indicated that the gradient distribution of
HHMP-Si-CC in the photopolymerization system resulted in the
gradient distributions of degree of polymerization and molecular
weight of PMMA initiated by HHMP-Si-CC, which would influence
the thermostability and Tg of the polymer. The thermostability and
Tg of the PMMA were therefore measured by thermal gravimetric
analysis (TGA) and differential scanning calorimetry (DSC). It was
observed from Table 2 that T5% and Tmax of the PMMA initiated by
3.8. Surface morphology of gradient polymer films
Oxygen inhibition is a commonly encountered issue while per-
forming radical polymerizations. Many applications based on
photopolymerizations namely paints, coatings, adhesives, photoli-
thography, and dental resins are severely affected by oxygen
Fig. 6. Effect of the concentration of HHMP-Si-CC on the polymerization of TPGDA
(I ¼ 40 mW cmꢀ2).
Fig. 8. The HHMP and HHMP-Si-CC concentrations of top and bottom layer of the
system using absolute ethyl alcohol as a solvent with different standing time.

F. Sun et al. / Polymer 55 (2014) 3656e3665
3663
Fig. 9. (a) XPS survey spectrum of the polymer (PTPGDA) initiated by HHMP-Si-CC
based on TPGDA; (b) Silicone content and atomic ratio of Si to C in each layer of the
polymer.
inhibition. Currently, strategies to reduce oxygen inhibition are
abundant, which include physical approaches such as nitrogen
inerting, lamination and use of higher light intensity, as well as
chemical strategies involving chemical additives [19e22], cationic
polymerization [23], highly branched or multifunctional mono-
mers, and acrylates with inherent reduced oxygen inhibition
[24e25]. Among the chemical additive strategies, the most obvious
method is to use a higher concentration of photoinitiator [26e27],
which provides a higher concentration of initiating radicals. While
the formed initiating radicals react more rapidly with molecular
oxygen than with acrylate monomers, the concentration of oxygen
within the system is initially decreased to a level that propagation
becomes preferred [28]. The HHMP-Si-CC with self-floating ability
can spontaneously enrich at the surface of the polymerization
system, leading to a higher concentration of HHMP-Si-CC at the
surface thereby decreasing oxygen inhibition. Here, we investi-
gated the influence of oxygen inhibition in the photo-
polymerization systems initiated by HHMP and HHMP-Si-CC
(Fig. 12).
Fig. 10. Raman spectra of PMMA obtained by using HHMP (a) and HHMP-Si-CC (b) as
photoinitiator (1.0 ꢂ 10ꢀ3 mol Lꢀ1).
concentration of HHMP-Si-CC, and almost disappeared when the
concentration of HHMP-Si-CC reached to 1.0 ꢂ 10ꢀ2 mol Lꢀ1, indi-
cating that HHMP-Si-CC can reduce the oxygen inhibition through
the enrichment on the surface of the photopolymerization system
caused by the good self-floating ability.
3.9. Migration of photolysis fragments of photoinitiator from the
cured material
To overcome the migration of photolysis fragments of the
photoinitiator from UV-curable materials, the polymerizable
groups were introduced into the HHMP-Si-CC molecule. The
When a liquid film of UV curable resin coated on a substrate is
exposed to UV light in the presence of oxygen, a thin liquid layer of
monomers remains uncured on the surface due to the quenching of
free radicals by oxygen [29]. The uncured liquid layer could spon-
taneously swell the underlying substrate-constrained gradient
polymer film, leading to in-plane stresses that generate surface
wrinkle patterns [30]. Remarkably, the photopolymerization sys-
tem initiated by HHMP in the presence of oxygen appeared a strong
oxygen inhibition, while the oxygen inhibition was reduced to a
certain extent by using the photoinitiator HHMP-Si-CC with self-
floating ability (Fig. 12). Besides, it was observed from Fig. 13 that
the wrinkling gradually became weak with the increase of the
Table 1
DBC of PMMA obtained by using HHMP and HHMP-Si-CC as photoinitiators
(1.0 ꢂ 10ꢀ3 mol Lꢀ1).
Photoinitiators
HHMP
Segment
DBC(%)
Top layer
97.23
97.43
98.02
97.33
99.23
98.98
97.25
95.62
Second layer
Third layer
Bottom layer
Top layer
Second layer
Third layer
Bottom layer
HHMP-CC

3664
F. Sun et al. / Polymer 55 (2014) 3656e3665
Fig. 14. HPLC chromatograms of extraction solutions of HHMP-Si-CC and HHMP.
Fig. 11. Molecular weight of PMMA obtained by using HHMP and HHMP-Si-CC as
photoinitiator (1.0 ꢂ 10ꢀ3 mol Lꢀ1).
photolysis fragments of HHMP-Si-CC can be polymerized into the
polymer backbone through the carbonecarbon double bonds, thus,
the migration of photolysis fragments from the cured material
would be mitigated.
Table 2
Here we investigated the migration of the photolysis fragments
of the HHMP-Si-CC from UV-curable films by HPLC. The HPLC
chromatograms of the extraction solution of the curable films
initiated by HHMP-Si-CC and HHMP respectively (Fig. 14) showed
that there were one intense peak at around 2.0 min and two small
peaks at around 1.7 min and 10.5 min. The intense peak was
assigned to the unreacted monomer TPGDA, while the small peak at
around 10.5 min was assigned to the impurity in the solvent, the
other small peak at around 1.7 min was the peak of photoinitiator
fragments. Remarkably, the peaks of photoinitiator fragments and
unreacted monomer TPGDA from the extraction solution of HHMP-
Si-CC were obviously weaker than those of HHMP, demonstrating
that HHMP-Si-CC with carbonecarbon double bonds prevented the
migration of photolysis fragments from the cured material to some
extent, and was an efficient photoinitiator.
Thermal decomposition parameters and Tg of each layer of PMMA prepared by using
HHMP and HHMP-Si-CC as a photoinitiator.
Photoinitiators
HHMP
Segments
T5%
/
ꢁC
Tmax
/
ꢁC
Tg/ ꢁC
Top layer
188.2
188.5
189.5
188.8
158.4
188.2
191.5
237.7
384.2
382.6
382.5
386.8
370.4
387.2
390.5
395.7
111.4
115.7
115.5
113.6
118.2
116.8
115.9
111.2
Second layer
Third layer
Bottom layer
Top layer
Second layer
Third layer
Bottom layer
HHMP-Si-CC
Note: T5% is the temperature of 5% weight loss. Tmax is the peak temperature at
maximum weight loss rate.
Fig. 12. SEM images of the polymer films (PTPGDA) initiated by HHMP and HHMP-Si-CC with the same molarity (1.0 ꢂ 10ꢀ2 mol Lꢀ1).
Fig. 13. SEM images of the polymer films (PTPGDA) initiated by different concentration of HHMP-Si-CC ((a) 4.0 ꢂ 10ꢀ3 mol Lꢀ1 (b) 8.0 ꢂ 10ꢀ3 mol Lꢀ1 (c) 1.0 ꢂ 10ꢀ2 mol Lꢀ1)).

F. Sun et al. / Polymer 55 (2014) 3656e3665
3665
4. Conclusions
References
[1] Zhang Y. J Eur Ceram Soc 2012;32:2623e32.
[2] Wen B, Wu G, Yu J. Polymer 2004;45:3359e65.
[3] Bhat RR, Chaney BN, Rowley J, Liebmann-Vinson A, Genzer J. Adv Mater
2005;17:2802e7.
[4] Kim J, Mok MM, Sandoval RW, Woo DJ, Torkelson JM. Macromolecules
2006;39:6152e60.
[5] Gritsenko KP, Krasovsky AM. Chem Rev 2003;103:3607e49.
[6] Cui Y, Yang J, Zeng Z, Chen Y. Eur Polym J 2007;43:3912e22.
[7] Yao D, Zhang W, Zhou JG. Biomacromolecules 2009;10:1282e6.
[8] Jiang D, Huang X, Qiu F, Luo C, Huang LL. Macromolecules 2010;43:71e6.
[9] Song C, Xu Z, Li J. Mater Des 2007;28:1012e5.
[10] Zhang N, Li M, Nie J, Sun F. J Mater Chem 2012;22:9166e72.
[11] Liska R. J Polym Sci Part A: Polym Chem 2002;40:1504e18.
[12] Teshima W, Nomura Y, Tanaka N, Urabe H, Okazaki M, Nahara Y. Biomaterials
2003;24:2097e103.
[13] Tar H, Esen DS, Aydin M, Ley C, Arsu N. Macromolecules 2013;46:
3266e72.
[14] Kitano H, Ramachandran K, Bowden N, Scranton AJ. Appl Polym Sci 2013;128:
611e8.
In this paper, a novel photocleavage polysiloxane macromolec-
ular photoinitiator with polymerizable group (HHMP-Si-CC) was
synthesized, and its structure was confirmed by 1H NMR, FTIR and
GPC. The UV absorption spectra showed that HHMP-Si-CC had a
maximum absorption peak at 275 nm and its intensity decreased
with the increase of the irradiation time. The kinetics of photo-
polymerization and SEM images of curable films initiated by
HHMP-Si-CC indicated that HHMP-Si-CC efficiently initiated pho-
topolymerization and mitigated the oxygen inhibition in radical
photopolymerization to a certain degree.
The UV absorption spectra and XPS proved that the HHMP-Si-CC
had relatively good self-floating ability and spontaneously formed a
gradient distribution of concentration in the polymerization system
after standing for a designated time. The polymer initiated by
HHMP-Si-CC presented a gradient change in the degree of poly-
merization, molecular weight, thermostability and Tg, indicating
that gradient polymeric material was obtained by using the pho-
toinitiator HHMP-Si-CC under irradiation of UV light. More
importantly, HPLC chromatograms demonstrated that the migra-
tion of photolysis fragments of HHMP-Si-CC with polymerizable
groups toward the cured material surface was mitigated signifi-
cantly. The HHMP-Si-CC should have potential applications in
preparing more environmentally friendly gradient materials.
[15] Wang H, Wei J, Jiang X, Yin J. J Polym Sci Part A: Polym Chem 2006;44:
3738e50.
[16] Yagci Y, Jockusch S, Turro NJ. Macromolecules 2010;43:6245e60.
[17] Hou G, Shi S, Liu S. Polym J 2008;40:228e32.
[18] Fu H, Zhang S, Xu T, Zhu Y, Chen J. Environ Sci Technol 2008;42:2085e91.
[19] Ibrahim A, Stefano LD, Tarzi O, Tar H, Ley C, Allonas X. Photochem Photobiol
2013;89:1283e90.
ꢀ
ꢁ
ꢀ
[20] Tejkl M, Valis J, Kaplanova M, Jasúrek B, Syrový T. Prog Org Coat 2012;74:
215e20.
[21] Roper TM, Kwee T, Lee TY, Guymon CA, Hoyle CE. Polymer 2004;45:
2921e9.
[22] Versace DL, Dalmas F, Fouassier JP, Lalevee J, Dichloride Z. Macro Lett 2013;2:
341e5.
[23] Crivello JV, Ma J, Jiang F, Hua H, Ahn J, Ortiz RA. Macromol Symp 2004;215:
165e78.
Acknowledgements
[24] Cai Y, Jessop JLP. Polymer 2006;47:6560e6.
[25] Decker C, Viet TNT, Decker D, Weber-Koehl E. Polymer 2001;42:5531e41.
[26] Hageman HJ. Prog Org Coat 1985;13:123e50.
[27] Hageman HJ, Jansen LGJ. Makromol Chem 1988;189:2781e95.
[28] Ligon SC, Husar B, Wutzel H, Holman R, Liska R. Chem Rev 2014;114:
557e89.
The financial support from National Natural Science Foundation
of China (Grant No. 51273014) is gratefully acknowledged.
Appendix A. Supplementary data
ꢁ
[29] Lalevee J, Fouassier JP. Polym Chem 2011;2:1107e13.
[30] Chandra D, Crosby AJ. Adv Mater 2011;23:3441e5.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.polymer.2014.06.040.