Dual roles for promoting monomers to polymers: A conjugated sulfonium salt photoacid generator as photoinitiator and photosensitizer in cationic photopolymerization

Article abstract of DOI:10.1002/pola.28154

An efficient strategy for comprehensive utilization of the conjugated sulfonium salt photoacid generator (PAG), namely, 3-{4-[4-(4-N,N′-diphenylamino)-styryl]phenyl}phenyl dimethyl sulfonium hexafluoroantimonate, was developed through photoinitiated cationic photopolymerization (CP) of epoxides and vinyl ether upon exposure to near-UV and visible light-emitting diodes (LEDs; e.g., 365, 385, 405, and 425 nm). Photochemical mechanisms were investigated by UV–vis spectra, molecular orbital calculations, fluorescence, cyclic voltammetry, and electron spin resonance spin-trapping analyses. Compared with commercial PAGs, the prepared conjugated sulfonium salt generated H⁺, which can be used as photoinitiator. Moreover, the fluorescent byproducts from photodecomposition can be used as photosensitizer of commercial iodonium salt in the photoinitiating systems of CP. These novel D-π-A type sulfonium-based photoinitiating systems are efficient (epoxide conversion = 85–90% and vinyl conversion >90%; LEDs upon exposure to 365–425 nm) even in low-concentration initiators (1%, w/w) and low curing light intensities (10–40 mW cm^?2).

Full text of DOI:10.1002/pola.28154

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Dual Roles for Promoting Monomers to Polymers: A Conjugated
Sulfonium Salt Photoacid Generator as Photoinitiator
and Photosensitizer in Cationic Photopolymerization
Ming Jin,1 Xingyu Wu,1 Jean Pierre Malval,2 Decheng Wan,1 Hongting Pu1
1School of Materials Science and Engineering, Tongji University, 4800 Caoan Road, Shanghai 201804, People’s Republic of China
2
ꢀ
ꢀ
Institute De Science Des Materiaux De Mulhouse, UMR CNRS 7361, Universite De Haute-Alsace, 15 Rue Jean Starcky, Mulhouse
68057, France
Correspondence to: M. Jin (E-mail: mingjin@tongji.edu.cn)
Received 5 April 2016; accepted 25 April 2016; published online 00 Month 2016
DOI: 10.1002/pola.28154
ABSTRACT: An efficient strategy for comprehensive utilization of
the conjugated sulfonium salt photoacid generator (PAG), namely,
3-{4-[4-(4-N,N0-diphenylamino)-styryl]phenyl}phenyl dimethyl sul-
fonium hexafluoroantimonate, was developed through photoiniti-
ated cationic photopolymerization (CP) of epoxides and vinyl ether
upon exposure to near-UV and visible light-emitting diodes (LEDs;
e.g., 365, 385, 405, and 425 nm). Photochemical mechanisms were
investigated by UV–vis spectra, molecular orbital calculations, fluo-
rescence, cyclic voltammetry, and electron spin resonance spin-
trapping analyses. Compared with commercial PAGs, the prepared
conjugated sulfonium salt generated H1, which can be used as
photoinitiator. Moreover, the fluorescent byproducts from photo-
decomposition can be used as photosensitizer of commercial iodo-
nium salt in the photoinitiating systems of CP. These novel D-p-A
type sulfonium-based photoinitiating systems are efficient (epox-
ide conversion 5 85–90% and vinyl conversion >90%; LEDs upon
exposure to 365–425 nm) even in low-concentration initiators (1%,
w/w) and low curing light intensities (10–40 mW cm22).
2016
C
V
Wiley Periodicals, Inc. J. Polym. Sci., Part A: Polym. Chem. 2016,
00, 000–000
KEYWORDS: cationic photopolymerization; conjugated sulfonium
salt; near UV–visible LEDs; photochemistry; photoinitiator;
photopolymerization; photosensitizer; radical polymerization
involves the association of a long-wavelength active sensitizer
with commercial PAGs, such as onium salts. Acid generation
proceeds through photoinduced electron transfer (PET) from
the excited sensitizer to PAGs. However, the global kinetics of
this bimolecular method is intrinsically limited by diffusion.
Therefore, an integrated approach is proposed; this approach
is based on the molecular association of the acid generator
functionality into the structure of the low energy photon
active chromophore.8,9,20–23 Perry et al.20 used this strategy to
INTRODUCTION Photoacid generators (PAGs) constitute an
important class of cationic photoinitiators and strategically
used in photoresists.1 PAGs are widely applied in various
research domains, such as microelectronics,2 photosensitized
resins,3 3D microfabrication, and high-density optical data
storage.4,5 Diaryliodonium and triarylsulfonium salts are the
most important PAGs because of their thermal stability, solu-
bility in most cationic polymerized monomers, and efficiencies
in generating strong protic acids during UV irradiation.6,7 In
recent years, light-emitting diodes (LEDs), which are cheaper
and safer irradiation devices than UV lamps, have gained
increased research attention.8,9 New photoinitiating systems,
for example, curcumin as a multicolor photoinitiator,10 N-
vinylcarbazole as versatile photoinaddimer,11 hexaarylbiimida-
zole as visible light photoinitiator12 using LEDs as light source
gained good polymerization conversions. However, the size of
aryl groups in commercial PAGs is small, and their absorption
peaks are below 300 nm; hence, these materials are unsuitable
when long-wavelength UV and visible LED sources are
applied.13–15 Two approaches can be used to achieve specific
excitation of PAGs near UV- and visible-light ranges.7–9,16–23
The first approach is derived from sensitizing methods and
design
a sulfonium-based bis[(diarylamino)styryl]benzene
derivative, which exhibits maximum absorption at 390 nm and
high quantum yield for acid generation (ꢀ0.5). We fabricated
4-ethoxystilbene and N,N-diphenylamino-substituted PAGs by
using the D-p-A system, in which a methyl(p-cyanobenzyl)sul-
fonium was monosubstituted in either para- or meta-position,
and prolonged the absorption peak to approximately
400 nm.8,9,22,23 Our prepared PAGs exhibited effective acid
generation efficiency (AH1 5 0.3–0.6). However, the molecular
weights of PAGs with big conjugated systems decrease the
molar contents of generated H1 in the photoinitiating systems,
and the ultimate conversion of cationic monomers is approxi-
mately 50–70%.8,9
C
V
2016 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000
1

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
spectra are spectrally corrected, and fluorescence quantum
yields include the correction due to solvent refractive index
are determined relative to quinine bisulfate in 0.05 M sulfu-
ric acid (Afl 50.52).24 The fluorescence lifetimes are meas-
ured using a Nano LED emitting at 372 nm as an excitation
source with a Nano LED controller module, Fluorohub from
IBH, operating at 1 MHz. The detection is based on an
R928P type photomultiplier from Hamamatsu with high sen-
sitivity photon-counting mode. The decays are fitted with the
iterative reconvolution method on the basis of the Mar-
quardt–Levenberg algorithm. Such a reconvolution technique
allows an overall-time resolution down to 0.15 ns. The qual-
ity of the exponential fits is checked using the reduced v2
(ꢁ1.2).
SCHEME 1 Chemical structures of the target sulfonium salt (PI-
PAG), its precursor (Precursor), iodonium salt (Iod), and cati-
onic monomers (EPOX, CHO, and DVE-3).
In this paper, we developed a p-conjugated sulfonium salt
(PI-PAG, Scheme 1) characterized by a comprehensive utiliza-
tion of two mechanisms, for example, integrated and sensitiz-
ing methods. The fabricated PI-PAG has high quantum yield
of acid generation (FH1 5 0.44) upon exposure to near
UV–vis LEDs (365–425 nm). Photolysis byproducts, which
have strong fluorescence, can synthesize iodonium salts to
regenerate H1 through PET. H1 in the two steps can initiate
the simultaneous photopolymerization of epoxides and vinyl
ethers, which can greatly improve efficiency and energy con-
sumption. The stepwise photoacid generation mechanism
was studied using UV–vis spectra, fluorescent, cyclic voltam-
metry, and ESR spin-trapping analyses. The ability of the
photoinitiating systems to start polymerization reactions was
verified by real-time FTIR experiment.
Quantum yields for acid generation were measured under
irradiation at 385 nm using LED point curing (Uvata, Shang-
hai). The irradiated PAG dissolved in acetonitrile was previ-
ously N2-degassed. The progress of the photoreaction was
monitored via UV–vis absorption spectra. The absorbance at
the excitation wavelength was greater than 2.5 to assume a
total absorption of the incident photons. The dose rates
were maintained sufficiently small so that the changes of Ak
were lower than 10%. Rhodamine B (RhB) was used as a
sensor of photoacid generation.25 The acid generation in ace-
tonitrile was also evaluated from a calibration curve of RhB,
which was gradually protonated by the addition of
p-toluenesulfonic acid. The incident light intensity was meas-
ured by ferrioxalate actinometry.
EXPERIMENTAL
Theoretical Calculations
Materials
The theoretical calculations have been computed based on
density functional theory (DFT). The overall computation
strategy was defined as follows: After initial AM1 optimiza-
tion calculations (in ACN), subsequent optimization of geo-
metrical structures of the compounds were carried out using
the B3LYP/6-31G(d) level of calculation. All calculations have
been performed using GAUSSIAN 09 package.26 The molecu-
lar orbitals involved in these transitions could be extracted,
and the geometries were frequency checked.
The PAG and monomers investigated here are presented in
Scheme 1 and used with the best purity available. (3,4-Epox-
ycyclohexane)methyl 3,4-epoxycyclohexylcarboxylate (EPOX),
cyclohexene oxide (CHO), and triethyleneglycol divinyl ether
(DVE-3) were purchased from Sigma-Aldrich or J&K Chemi-
cal and used as received without further purification. All
chemicals for synthesis were purchased from Sinopharm
Chemical Reagent; TCI; or J&K Chemical, and they were used
without further purification unless otherwise specified. All
solvents employed for photophysical measurements were
J&K Chemical spectroscopic grade.
Redox Potentials
The cyclic voltammetry experiments27 (using a computer-
controlled Radiometer Voltalab 6 potentiostat with a three-
electrode single-compartment cell, in which the working
electrode is a platinum disk, and a saturated calomel elec-
trode (SCE) that is used as a reference that placed in a sepa-
General Instruments
1H NMR and 13C NMR measurements were recorded with a
Bruker 400M NMR spectrometer and chemical shifts were
reported in parts per million (ppm) downfield from the
Me4Si resonance, which was used as the internal standard
when recording NMR spectra. Elemental analysis was per-
formed by Elementar Vario El III (Germany). Mass spectra
were recorded on a Micromass GCTTM.
rate compartment with
a salt bridge containing the
supporting electrolyte and saturated KCl performed at
300 K, in acetonitrile with a constant concentration (0.1 M)
of n-Bu4NBF4. Ferrocene (Fc) is used as an internal
reference.
Photochemical and Photophysical Experiments
ESR Spin Trapping Experiment
UV–vis spectra were recorded on a Mapada UV-6300 spectro-
photometer. The photodecomposition processes in a solution
of acetonitrile were studied by UV–vis spectra under an LED
irradiation of 385 nm. Steady-state fluorescence spectra are
collected from a Hitachi-2700 spectrofluorometer. Emission
Electon spinning resonance spin trapping (ESR-ST) experi-
ments carried out using a Bruker EMX-plus spectrometer
(X-band). The radicals were generated at room temperature
upon 365-nm LED exposure under Ar and trapped by phe-
nyl-N-tert-butylnitrone (PBN) in tert-butylbenene.28
2
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
13C NMR (101 MHz, CDCl3, d, ppm): d 147.57, 147.49,
141.46, 139.50, 139.06, 137.03, 131.45, 129.36, 129.26,
128.45, 127.46, 127.41, 126.79, 126.44, 125.40, 125.12,
124.58, 123.84, 123.58, 123.13, 16.34. MS (m/z): calcd. for
C33H27NS, 469.19; found, 470.19 [M 1 H1]. Anal. calcd for
C33H27NS: C, 84.40; H, 5.79; N, 2.98. Found: C, 84.31; H,
5.74; N, 2.90.
Preparation of 3-(4-(4-(4-N,N0-Diphenylamino)-
Styryl)Phenyl)Phenyl Dimethyl Sulfonium
Hexafluoroantimonate (PI-PAG)
Precursor (197.7 mg, 0.4 mmol), Cs2CO3 (274.3 mg, 0.8
mmol) was added to a 100-mL flask along with 20.0 mL of
dry dichloromethane and the mixture was stirred under 278
8C for 30 min. Then methyl trifluoromethanesulfonate
(76.0 mg, 0.5 mmol) was syringed into the solution.34 After
stirring at this temperature for 2 h, the mixture was warmed
to room temperature and stirred for another 48 h. Then, col-
umn chromatography was used for separation and purifica-
tion (dichloromethane: methanol 1: 5, v/v, 500 mL) to
obtain yellow product (sulfomium trifluoromethanesulfo-
nate). Then, anion exchange of CF3SO23 by SbF62 was per-
formed using the Zhou et al. method.35 A fresh 20-mL
portion of NaSbF6 (1.14 g, 4.40 mmol) in water was added
SCHEME 2 Synthetic routes of the PAG. (a) NBS, CCl4, BPO;
(b) P(OEt)3, 150 8C, 3 h; (c) potassium tert-butoxide, THF, 0 8C;
(d) Pd(PPh3)4, K2CO3, toluene/ethanol/H2O; (e) CF3SO3CH3,
Cs2CO3, CH2Cl2, RT; and (f) NaSbF6, water, and acetone.
Photopolymerization Experiments
The photopolymerization is monitored in situ by Fourier
transform real-time infrared spectroscopy (FT-RTIR) with a
Thermo-Nicolet iS5 instrument IR spectrometer. The photo-
sensitive formulations are deposited on a KBr pellet in lami-
nate (40 lm thick) for irradiation at different wavelength
LEDs (Uvata, Shanghai). The evolution of the epoxy group
content of EPOX, the epoxy group content of CHO, the double
bond (C 5 C) content of DVE-3 are continuously followed at
about 790, 780, and 1620 cm21, respectively.29,30
to
a
solution of sulfomium trifluoromethanesulfonate
(100 mg, 0.16 mmol) in 10 mL of acetone. The mixture was
stirred for 2 h at room temperature in the dark. The solid
was collected by filtration and redissolved in 10 mL of ace-
tone. The above anion exchange was repeated three times.
The resulting solid was washed three times with water and
ether. The product was purified by two precipitations from a
10-mL acetone solution through the addition of 50 mL of
diethyl ether. Yellow product was gained and the yield:
89.0%.
Synthesis
The synthetic route of the precursor and PAG are shown in
Scheme 2, and the details of synthesis and characterization
are provided below. Compound 1, 2, and 3 were prepared
following the procedure of literature.31,32
1H NMR (400 MHz, CDCl3, d, ppm): d 7.92 (s, 1H, PhH), 7.72
(t, J 5 7.3 Hz, 2H, PhH), 7.53 (m, 3H, PhH), 7.42 (d, J 5 8.0
Hz, 2H, PhH), 7.31–7.14 (m, 6H, PhH), 7.04 (d, J 5 7.9 Hz,
4H, PhH), 6.98 (m, 5H, PhH, CH 5 CH), 6.83 (d, J 5 16.3 Hz,
1H), 3.11 (s, 6H, CH3). 13C NMR (101 MHz, CDCl3, dppm): d
147.53, 147.39, 143.40, 138.05, 136.31, 132.58, 131.80,
130.90, 129.31, 129.11, 127.54, 127.45, 127.39, 127.05,
125.66, 125.36, 124.58, 123.23, 123.15, 28.39. MS (m/z):
calcd for C34H30NS1, 484.21; found, 484.21 [sulfonium ion].
Anal. calcd for C34H30F6NSSb: C, 56.68; H, 4.20; N, 1.94.
Found: C, 56.63; H, 4.14; N, 2.03.
Preparation of 4-(3-Methylthiophenyl)24-N,N0-
diphenylamino-Stilebene (Precursor)
A 250-mL three-necked, round-bottomed flask fitted with an
electromagnetic stirrer and a reflux condenser was charged
with 3 (1.1 g, 2.5 mmol), 3-(methylthio)phenylboronic acid
(420.1 mg, 2.5 mmol), potassium carbonate (1.0 g, 7.5
mmol), methyl trioctyl ammonium chloride 0.3 mL,
Pd(PPh3)4 (86.7 mg, 0.08 mmol), and toluene/ethanol/H2O
(v/v/v, 5:3:1, 40 mL: 24 mL: 8 mL).33 The resulting mixture
was heated to reflux with stirring under nitrogen at approxi-
mately 90 8C for 48 h. The mixture was extracted by
dichloromethane (50 mL 3 3). Then column chromatography
was used for separation and purification (dichloromethane:
petroleum ether 1: 5, v/v, 500 mL) to obtain yellow product
Precursor (1.1 g, 92.7% yield).
RESULTS AND DISCUSSION
Synthesis
The target PAG, named as PI-PAG, was prepared from
4-aldehyde-triphenylamine through four steps (Scheme 2)
because the triphenylamine group is a good substituent in
sulfonium salts.35 The Suzuki reaction was performed with
meta-(methylthio)-phenylboronic acid to produce precursor
(Precursor) in high yields (over 90%). Finally, alkylation of
the Precursor with CH3CF3SO3 generated a conjugated sulfo-
nium triflate salt; subsequently, non-nucleophilic anions,
1H NMR (400 MHz, CDCl3, d, ppm): d 7.57 (s, 4H, PhH), 7.49
(s, 1H, PhH), 7.44–7.33 (m, 4H, PhH), 7.31–7.21 (m, 5H,
PhH), 7.17–6.99 (m, 10H, PhH, CH 5 CH), 2.54 (s, 3H, CH3).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000
3

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 2 Representation of frontier molecular orbitals
involved in the lowest energy electronic transitions of Precur-
sor and PI-PAG. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
FIGURE 1 The UV–vis spectra of PI-PAG and Precursor in ace-
tonitrile. [Color figure can be viewed in the online issue, which
is available at wileyonlinelibrary.com.]
conjugated and sulfonium moieties, and an apparent p–r*
(sulfur-carbon bond, face to face) transition is involved in
the LUMO orbital (Fig. 2), thereby decreasing the molar
extinction coefficients of PI-PAG but increasing the electron
delocalization in the molecular structure; these changes
caused the spectra redshift relative to its precursor. Saeva
confirmed that the photoacid generation reaction of arylsul-
fonium salts may occur either by direct excitation of p–r*
(sulfur–carbon bond) transition or excitation of the p–p*
transition, followed by intramolecular electron transfer to r*
on the sulfur–carbon bond.34,37 In both cases, subsequent
homolytic cleavage of the S-C bond leads to the formation of
radical cation on the sulfur-containing fragment and a neu-
tral radical on the carbon fragment.
such as SbF–6, were exchanged to improve the polymerization
reactions.35 The successful ions exchange can be proved by
the elemental analysis due to the different molecular weight
of anions (SbF26 : 235.7 and CF3SO23 : 149.1 as shown in
Experimental section).
Absorption Properties of the Sulfonium Salt
and Precursor
The absorption spectra of sulfonium salt and its precursor
(PI-PAG and Precursor) in acetonitrile are shown in Figure 1.
The maxima for PI-PAG and Precursor are located at kmax 5
381 nm (e381 nm 5 22,500 M21 cm21) and kmax 5 374 nm
(e374 nm 5 35,000 M21 cm21), respectively (Table 1). The
spectrum of PI-PAG presents a slight redshift but a lower e
value compared with its precursor (Precursor). The excellent
absorption properties of the compounds are associated with
the large delocalization of the highest occupied molecular
orbital (HOMO) and the lowest unoccupied molecular orbital
(LUMO), which are involved in the lowest energy
transition.24,36
A possible photolysis mechanism is proposed as shown in
Scheme 3. The photogenerated proton can initiate the cati-
onic polymerization of epoxides and vinyl ether.
Cationic Photopolymerizations
Photopolymerization of EPOX
The photopolymerization of EPOX in the presence of PI-PAG
(1 wt %) in air were conducted through exposure to 365–
425 nm LEDs because the PI-PAG showed considerable
extinction coefficients within this range (Fig. 1 and Table 1).
PI-PAG can initiate the polymerizations of EPOX; the conver-
sion profile for the reaction and the conversion at t 5 5 min,
which is 52.3% at 365 nm LED, is depicted in Figure 3(a).
Improved polymerization profiles were obtained upon addi-
tion of commercial iodonium salt (Iod), which is used in
photosensitizing systems in visible-light photopolymeriza-
tion,30 to the PI-PAG system [Fig. 3(a)]. The conversions of
This result was further confirmed by DFT calculations, as
shown in Figure 2. The HOMOs of both PI-PAG and precursor
are mainly localized on the triphenylamine (donor) and stil-
bene (conjugated system) moieties; however, the LUMO of
Precursor is distributed throughout the conjugated system.
The p–p* transition of HOMO-LUMO orbitals explains the
high e values (Fig. 1). The LUMO of PI-PAG is located on the
TABLE 1 Light Absorption Properties of PI-PAG and Precursor and Their Molar Extinction Coefficients (e in M21 cm21
)
at the Maximum Emission Wavelengths of Different LEDs
a
b
b
b
b
1c
Compound
kmax (nm)a
emax
e365 nm
e385 nm
e405 nm
e425nm
FH
Precursor
PI-PAG
374
381
35,000
23,200
32,000
19,200
32,000
23,000
14,500
14,900
1,900
3,800
–
0.44
a
c
In acetonitrile, and the error is approximately 10%.
For different near UV/visible LEDs.
Measured at 385 nm, and the error is approximately 15%.
b
4
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
SCHEME 3 Proposed photolysis mechanism of sulfonium salt and photogenerated H1 and fluorescent byproducts. [Color figure
can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
the two component PI-PAG/Iod systems reach 82.0%, which
can be considered as excellent photoinitiating systems for
this viscous monomer.38 The satisfactory performance of
these two-component systems could be due to the contribu-
tion of a PET from the excited sensitizer to the Iod, as dis-
cussed in a previous report.19 According to the photolysis
mechanism shown in Scheme 3, the precursor (Precursor)
exists in the photoinitiating systems; thus, the possible pho-
tosensitization process was studied in the present work.
The related polymerization profiles are provided in Figure
3(a). Iod and Precursor could not initiate the conversion of
EPOX. Precursor/Iod (1%/3%, w/w) system can initiate the
polymerizations of EPOX, with a relative conversion of
approximately 30–45%. Thus, we could preliminary conclude
that the satisfactory performance of the PI-PAG/Iod system
could be due to further photosensitization of byproducts of
PI-PAG photolysis. The exposure wavelengths were red-
shifted to the visible range (425 nm), the dicomponent sys-
tems still exhibit good conversions compared with the PI-
PAG itself [Fig. 3(b)]; this result is highly dependent on the
photosensitizing process. Final tack free coatings are
obtained within 5 min of LED irradiation in different wave-
lengths. Here, SbF26 was selected as counterion and it proves
that it has better conversion of EPOX than PF26 as anion
even the commercial Iod with PF26 as anion (the data are not
shown).
Photopolymerization of CHO And DVE-3
These systems also initiated the cationic polymerizations of
other types of monomers, for example, CHO and DVE-3. In
this study, the conversion profiles of CHO and DVE-3 at
405 nm were selected as samples to compare their dynam-
ics; the results are shown in Figure 4. These photoinitiating
systems are very efficient, and 0.5 wt % of PI-PAG is suffi-
cient to convert the monomer to polymers at approximately
100 s. At other wavelengths, the conversions of CHO reach
85–95% (Table 2).
FIGURE 3 (a) Photopolymerization profiles of EPOX in air and
in the presence of Iod (3 wt %), Precursor (1 wt %), PI-PAG
(1 wt %), PI-PAG/Iod (1%/3%, w/w), and Precursor/Iod (1%/3%,
w/w); and (b) Photopolymerization conversions of EPOX in air
and in the presence of different initiating systems at different
LED irradiation (40 mW cm22). [Color figure can be viewed in
the online issue, which is available at wileyonlinelibrary.com.]
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000
5

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 4 Photopolymerization profiles of CHO and DVE-3
under air in the presence of PI-PAG (0.5 wt %) and PI-PAG/Iod
(0.5%/2%, w/w) at 405 nm LED irradiation. [Color figure can be
viewed in the online issue, which is available at wileyonlineli-
brary.com.]
FIGURE 5 UV–vis spectra of PI-PAG in acetonitrile under irradi-
ation of 385 nm LED (1 mW cm22) at different periods. [Color
figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
The detailed photolysis and photosensitization processes
were also studied to clarify the possible polymerization
mechanism.
Photochemical and Photophysical Properties
of PI-PAG And Precursor
Photodecomposition and Photoacid Generation
of PI-PAG And PI-PAG(Precursor)/Iod Systems
The typical evolution of the absorption spectra of PI-PAG in
acetonitrile upon irradiation by 385 nm LED at 1 mW cm22
is shown in Figure 5. With the appearance of a less intensive
band located at 356 nm, the band located at approximately
280 nm increases slightly, whereas the maximum wavelength
absorption band dramatically collapses. During photolysis,
H1 is produced and the quantum yield of acid generation of
PI-PAG is 0.44 at 385 nm LED irradiation (Table 1 and Fig.
6). Additionally, the photoacid generation of the PI-PAG/Iod
system [Fig. 6(a)] and Precursor/Iod systems [Fig. 6(b)] was
also studied and compared with that of the PI-PAG itself. The
absorption peaks of RhB at 555 nm rapidly increases during
TABLE 2 CHO and DVE-3 Conversions Obtained with
365–425 nm LED Illumination in the Presence of PI-PAG/Iod
(0.5%/2%, w/w) with Light Intensity 10 mW cm22
Monomer
LEDs (nm)
Conversion (%)
CHO
365
385
405
425
365
385
405
425
87.5
85.2
83.5
84.9
90.6
89.0
90.3
88.4
CHO
CHO
FIGURE 6 (a) Photoacid generation curves of the PI-PAG/Iod
(1:2 mol/mol) mixture upon the LED exposure at 365 nm; UV–
vis spectra recorded at different irradiation times; and
(b) absorbance at 555 nm of RdB in different photoacid gener-
ating systems. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
CHO
DVE-3
DVE-3
DVE-3
DVE-3
6
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Precursor were further investigated. Photopolymerization
with the Precursor proved that it can be used as a cationic
photoinitiator with iodonium or sulfonium salts, wherein the
efficient production of reactive species (radicals and cations)
play a very important role. The well-known mechanism of
dye-sensitized photodecomposition of iodonium salts6 is pre-
sented in reactions 1–6 (r1–r6). The key step is PET between
dye/iodonium photoredox pair. The PET from the excited
dye (Dye*) to iodonium (Ph2I1) leads to the dye radical cat-
ion (Dye•1) and neutral radical (Ph•). The rapid decomposi-
tion of the Iod retards reverse electron transfer and renders
the overall process as irreversible. The radical cation (Dye•1
)
may allow the CP of monomers of epoxides and vinyl ethers.
Simultaneously, the hydrogen abstraction reaction between
the radical cation and solvents or monomers generates
strong protic acid, which can also initiate CP.
FIGURE 7 ESR spectra of radicals generated in (a) PI-PAG,
(b) PI-PAG/Iod, and (c) Precursor/Iod and trapped by PBN in
tert-butylbenzene: spectra. PBN/free radical adducts: (a) aN 5
14.7 G; aH 5 3.13 G; (b) aN 5 14.4 G; aH 5 2.4 G; and (c) aN 5
14.4 G; aH 5 2.4 G. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
Dye ! Dye ꢂ ðhvÞ
(r )
1
Dye ꢂ 1 Ph2I1 ! Dye•11 Ph•1 Ph2I
M2H 1 Dye•1 ! Dye12H 1 M•
Dye12H ! Dye 1 H1
(r )
2
(r )
3
(r )
4
the first 5 s and then increases slowly in the PI-PAG 1 2 eq.
Iod systems [Fig. 6(a)]. These results suggest that the photo-
acid generation could be divided into two stages [Fig. 6(b)].
The photolysis of PI-PAG is completed within 5 s under irra-
diation, and the byproducts are further sensitized the Iod to
gradually produce H1. Given that Precursor was used as sen-
sitizer, a similar data could be obtained in the same experi-
mental conditions. The photosensitization starts in the first
seconds; the enlarged data are observed starting at 5 s. This
finding suggests that H1 is produced during the photolysis
of the PI-PAG 1 Iod system.
M2H 1 Ph• ! Ph2H 1 M•
(r )
5
M•1 Ph2I1 ! M11 Ph2I 1 Ph•
(r )
6
In this PET process, the Gibbs free energy that changes the
electron transfer (DGet) from PIs to Iod (a kind of iodonium
salt) should be obtained to determine the potential applic-
ability of these dyes as electron donor in the excited
state.19,41 Free energy can be estimated according to Rehm–
Weller equation (eq 1)41 if we assume that this PET process
proceeds from the singlet excited state. The electron transfer
from the dyes to Iod is thermodynamically favorable if the
free energy (DGet) is negative.
EST-ST
The photochemical reactions were investigated through ESR-
ST experiments, and the corresponding signals are shown in
Figure 7. When PI-PAG was photodecomposed and the free
radicals were trapped by PBN [Fig. 7(a)], the hyperfine cou-
pling (hfc) constants of the radical adduct ESR of PI-PAG are
aN 5 14.7 G and aH 5 3.13 G, which is due to CH•3 generation
during the photolysis;28 these findings are strong evidence for
the photolysis mechanism of PI-PAG, as shown in Scheme 3.
When Iod was added to the ESR-ST experiments [Fig. 7(b)],
the hfc changed to aN 5 14.4 G and aH 5 2.4 G, which is similar
to the mixture of Precursor 1 Iod: aN 5 14.4 G and aH 5 2.4 G
[Fig. 7(c)]. These hfc values suggested that the trapped radi-
cals are phenyl radicals generated from Iod, as discussed by
Lalevee et al.39 The PET process occurred between the sensi-
tizers and Iod.
À
Á
Â
À
Á
Ã
DGet kJ mol21 5 97 Eox S=Sꢃ1 – EredðAꢃ–=AÞ 2 E00ðSÞ
2 Z1Z2=er12
(1)
In this equation, Eox and Ered correspond to the oxidation
potential of the Precursor [Eox 5 0.72 V (vs. SCE)] and the
reduction potential of Iod [Ered 5 20.2 V (vs. SCE) in ref.
43], respectively. E00 is the energy of the singlet excited
state determined by the normalized absorption and emission
spectra (E00 5 2.94 eV), and Z1Z2/er12 is the Coulombic
energy characterizing the interaction of the generated radi-
cal ion pairs with charges Z1 and Z2 in a medium of dielec-
tric constant e for distance r12. Given that the neutral radical
of the Iod is formed during PET process (Z 5 0), Z1Z2/er12
could be omitted in the calculation.24,44 The calculated DGet
is 2–194.20 kJ mol21, indicating that the combination of
Photochemical Properties of Precursor
Marder and Saeva et al. proved that the byproducts of PAG
photodecomposition are the main precursor and free radical
adducts.34,35 Our study on the excited state dynamics of a D-
p-A type sulfonium-based alkoxystilbene PAG also led to the
same conclusions.40 Thus, the relative spectral properties of
Precursor/Iod system possess
a
high driving force
(2DGet > 180 kJ mol21) upon light exposure. This finding
confirms that the PET process easily occurs through the sin-
glet excited state.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000
7

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
accelerating photobleaching of the Precursor in the presence
of Iod under irradiation of 385 nm LED further proves the
photochemical process (Fig. 9). The photobleaching could
also be used for other applications, for example, anti-
yellowing.
Photochemical and photophysical properties of PI-PAG and
precursor explain the occurrence of polymerization.
CONCLUSIONS
In conclusion, a new D-p-A type sulfonium salts photoacid
generator (PI-PAG) bearing the non-nucleophilic counterion
SbF–6 were used as cationic photoinitiators. The key features
of this sulfonium salt are as follows: (i) it produces H1 effi-
ciently upon exposure to 365–425 nm at very low light
intensity and (ii) the resulting fluorescent byproducts (pre-
cursor and H-abstraction products) can sensitize the Iod in
the cationic polymerization systems. Furthermore, they can
also be used as two-component photoinitiators in different
types of cationic monomers. Photochemically generated H1
was useful for acid-activated material processing and fabrica-
tion in the solid phase. This approach suggested a new
design strategy for the synthesis of p-conjugated sulfonium
salts, which exhibits dual roles when applied to photocuring.
Such PI-PAG can also be used as the sensitizer of sulfonium
salts, such as commercial triphenylsulfonium salt. Further
studies on the topic are in progress.
FIGURE 8 Fluorescence quenching of Precursor by Iod in ACN.
Inset: Stern-Volmer plot of fluorescence quenching. [Color fig-
ure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
The intermolecular electron transfer from Precursor to Iod
was also verified by fluorescence quenching measurements.
Precursor show good fluorescent emission, wherein the fluo-
rescence quantum yields are Afl 5 0.83 in ACN. The presence
of Iod leads to fluorescence quenching, as illustrated in Fig-
ure 8. The absence of any new peaks in the emission spectra
excludes any exciplex formation.
The singlet quenching constant (kq) is calculated from
eq 219 by using the fluorescence lifetime of the dye
(s0 5 2.5 ns) in the absence of a quencher (Table 1).
ACKNOWLEDGMENTS
I0=I5kqs0½Iodꢄ11
(2)
This work was supported by the National Natural Science Foun-
dation of China (51173134, 51573139); Fundamental Research
Funds for the Central Universities and the Open Measuring
Fund for Large Instrument and Equipment (0002015033) of
Tongji University.
For this system, the singlet quenching constant (kq) shows
that the value (5.35 3 1010 M21 s21) is close to the diffu-
sion control limit.44 These results confirm that fluorescence
quenching proceeds via the PET process.45 In addition, the
REFERENCES AND NOTES
1 E. Reichmanis, L. F. Thompson, In Polymers in Microlithogra-
phy. Am. Chem. Soc. 1989. p. 1–24.
2 M. Shirai, M. Tsunooka, Prog. Polym. Sci. 1996, 21–45.
3 J. P. Fouassier, Ed. Photoinitiator, Photopolymerization and
Photocuring: Fundamentals and Applications; Hanser Publish-
ers: New York, 1995.
4 S. Davidson, Ed. Exploring the Science, Technology and
Application of UV and EB Curing; Sita Technolgy Ltd: London,
1999.
5 J. P. Fouassier, J. Lalevee, Eds. Photoinitiator for Polymer
Synthesis-Scope, Reactivity, and Efficiency; Wiley: Weinheim,
2012.
6 J. V. Crivello, Ed. Photoinitiators for Free Radical, Cationic
and Anionic Photopolymerization; Wiley: Chichester, 1998.
FIGURE 9 Steady-state photolysis of Precursor and Precursor/
Iod mixture upon the LED exposure at 385 nm; A/A0 recorded
at different irradiation times. [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]
7 J. V. Crivello, M. Sangermano, J. Polym. Sci. Part A: Polym.
Chem. 2001, 39, 343–356.
8 M. Jin, X. Wu, J. Xie, J. P. Malval, D. Wan, RSC Adv. 2015, 5,
55340–55347.
8
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
9 M. Jin, H. Hong, J. Xie, J. P. Malval, A. Spangenberg, O.
Soppera, D. Wan, H. Pu, D. L. Versace, T. Leclerc, P. Baldeck,
O. Poizat, S. Knopf, Polym. Chem. 2014, 5, 4747–4755.
27 P. F. Hapiot, L. D. Kispert, V. V. Konovalov, J. M. Saveant, J.
Am. Chem. Soc. 2001, 123, 6669–6677.
ꢀ
28 A. Criqui, J. Lalevee, X. Allonas, J. P. Fouassier, Macromol.
10 J. Zhao, J. Lalevee, H. Lu, R. MacQueen, S. H. Kable, T. W.
Schmidt, M. H. Stenzel, P. Xiao, Polym. Chem. 2015, 6, 5053–
5061.
Chem. Phys. 2008, 209, 2223–2231.
29 J. Zhang, M. Frigoli, F. Dumur, P. Xiao, L. Ronchi, B. Graff,
F. Morlet-Savary, J. P. Fouassier, D. Gigmes, J. Lalevee, Macro-
molecules 2014, 47, 2811–2819.
11 P. Xiao, J. Lalevee, J. Zhao, M. H. Stenzel, Macromol. Rapid
Commun. 2015, 36, 1675–1680.
30 M. A. Tehfe, F. Dumur, E. Contal, B. Graff, F. Morlet-Savary,
D. Gigmes, J. P. Fouassier, J. Lalevee, Polym. Chem. 2013, 4,
1625–1634.
12 D. Ahn, S. S. Sathe, B. H. Clarkson, T. F. Scott, Dent. Mater.
2015, 31, 1075–1089.
13 G. Liu, X. Zhu, B. Xu, X. Qian, G. Song, J. Nie, J. Appl.
Polym. Sci. 2013, 130, 3698–3703.
31 M. Urbani, K. Ohkubo, D. M. S. Islam, S. Fukuzumi, F.
Langa, Chem. Eur. J. 2012, 18, 7473–7485.
14 M. A. Tehfe, F. Dumur, B. Graff, F. Morlet-Savary, J. P.
Fouassier, D. Gigmes, J. Lalevee, Macromolecules 2012, 45,
8639–8647.
32 Y. X. Yan, X. T. Tao, Y. H. Sun, W. T. Yu, G. B. Xu, C. K.
Wang, H. P. Zhao, J. X. Yang, X. Q. Yu, M. Zhao, M. H. Jiang,
Bull. Chem. Soc. Jpn. 2005, 78, 300–306.
15 P. Xiao, J. Zhang, F. Dumur, M. A. Tehfe, F. Morlet-Savary,
B. Graff, D. Gigmes, J. P. Fouassier, J. Lalevee, Prog. Polym.
Sci. 2015, 41, 32–66.
33 J. Wang, C. J. Yang, H. N. Peng, Y. S. Deng, X. C. Gao,
Synth. Commun. 2011, 41, 832–840.
34 F. D. Saeva, D. T. Breslin, H. R. Luss, J. Am. Chem. Soc.
1991, 113, 5333–5337.
16 Y. Y. Durmaz, N. Moszner, Y. Yagci, Macromolecules 2008,
41, 6714–6718.
35 W. H. Zhou, S. M. Kuebler, D. Carrig, J. W. Perry, S. R.
Marder, J. Am. Chem. Soc. 2002, 124, 1897–1901.
17 J. Lalevee, M. El-Roz, X. Allonas, J. P. Fouassier, J. Polym.
Sci. Part A: Polym. Chem. 2008, 46, 2008–2014.
36 M. Jin, J. Xie, J. P. Malval, A. Spangenberg, O. Soppera, D.
L. Versace, T. Leclerc, H. Pan, D. Wan, H. Pu, P. Baldeck, O.
Poizat, S. Knopf, J. Mater. Chem. C 2014, 2, 7201–7215.
18 R. Podsiadly, A. Maruszewska, R. Michalski, A. Marcinek, J.
Kolinska, Dyes Pigments 2012, 95, 252–259.
19 R. Podsiadły, K. Podemska, A. M. Szymczak, Dyes Pigments
2011, 91, 422–426.
37 X. Z. Wang, F. D. Saeva, J. A. Kampmeier, J. Am. Chem.
Soc. 1999, 121, 4364–4368.
20 W. H. Zhou, S. M. Kuebler, K. L. Braun, T. Y. Yu, J. K.
Cammack, C. K. Ober, J. W. Perry, S. R. Marder, Science 2002,
296, 1106–1109.
ꢀ
38 M. A. Tehfe, J. Lalevee, F. Morlet-Savary, B. Graff, N.
Blanchard, J. P. Fouassier, ACS Macro Lett. 2011, 1, 198–203.
39 J. Lalevee, H. Mokbel, J. P. Fouassier, Molecules 2015, 20,
7201–7221.
21 C. O. Yanez, C. D. Andrade, K. D. Belfield, Chem. Commun.
2009, 827–829.
40 V. De Waele, M. Hamm, T. Vergote, H. Chaumeil, M. Jin, J.
P. Malval, P. Baldeck, O. Poizat, Chem. Mater. 2015, 27,
1684–1691.
22 R. Xia, M. Jin, D. Wan, H. Pan, H. Pu, Prog. Chem. 2011, 23,
1854–1861.
23 M. Jin, H. Xu, H. Hong, J. P. Malval, Y. Zhang, A. Ren, D.
Wan, H. Pu, Chem. Commun. 2013, 49, 8480–8482.
41 K. Podemska, R. Podsiadły, A. M. Szymczak, E.
ꢀ
ꢀ
Chrzescijanska, J. Sokołowska, Dyes Pigments 2012, 95, 74–78.
24 R. Xia, J. P. Malval, M. Jin, A. Spangenberg, D. Wan, H. Pu,
T. Vergote, F. Morlet-Savary, H. Chaumeil, P. Baldeck, O.
Poizat, O. Soppera, Chem. Mater. 2012, 24, 237–244.
42 G. Yilmaz, S. Beyazit, Y. Yagci, J. Polym. Sci. Part A: Polym.
Chem. 2011, 49, 1591–1596.
43 R. Podsiadly, R. Strzelczyk, Dyes Pigments 2013, 97,
25 G. Pohlers, J. C. Scaiano, R. Sinta, Chem. Mater. 1997, 9,
462–468.
3222–3230.
44 B. Aydogan, G. E. Gunbas, A. Durmus, L. Toppare, Y. Yagci,
Macromolecules 2010, 43, 101–106.
26 M. J. Frisch, Ed. Gaussian 09, Revision B.01.; Gaussian, Inc.:
Wallingford, CT. Wallingford CT2009.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2016, 00, 000–000
9