The 2-benzoyl xanthone/triethylamine system as a type II photoinitiator: A laser flash photolysis and computational study

Article abstract of DOI:10.1016/j.jphotochem.2011.02.028

2-Benzoylxanthone (BzX) was synthesized, characterized and used as type II photoinitiator (PI) in combination with triethylamine for the polymerization of methylmethacrylate (MMA). The photophysical/photochemical behaviour of the photoinitiator, the invol

Full text of DOI:10.1016/j.jphotochem.2011.02.028

Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
The 2-benzoyl xanthone/triethylamine system as a type II photoinitiator: A laser
flash photolysis and computational study [1]
Xenophon Asvosa, Michael G. Siskosa,∗, Antonios K. Zarkadisa, Ralf Hermannb, Ortwin Bredeb
a Department of Chemistry, University of Ioannina, 45110 Ioannina, Greece
b Wilhelm-Ostwald-Institute for Physical and Theoretical Chemistry, Faculty of Chemistry and Mineralogy, University of Leipzig, Permoserstr. 15, 04318 Leipzig, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
2-Benzoylxanthone (BzX) was synthesized, characterized and used as type II photoinitiator (PI) in
combination with triethylamine for the polymerization of methylmethacrylate (MMA). The photophys-
ical/photochemical behaviour of the photoinitiator, the involved excited state and the reaction with the
amine co-initiator was studied by means of absorption and nanosecond time-resolved absorption spec-
troscopy. Upon irradiation with 266 or 355 nm laser light, the triplet state 3BzX* (ꢀmax = 355 nm and
530 nm) was generated as the only transient (lifetime of 22.7 ␮s) in nitrogen saturated MeCN solution.
3BzX* was confirmed through quenching experiments with oxygen, 2-methylbutadiene, perylene and
MMA and spectral similarity to benzophenone triplet (3BP*). The quantum yield of its formation (˚T = 0.8)
and the molar absorption coefficient (ε = 7500 L mol−1 cm−1) was measured in MeCN. The triplet was pho-
toreduced by triethylamine (TEA) via photoinduced electron/proton transfer giving the corresponding
ketyl and ␣-amino ethyl radical (•CHMe–NEt2). From the reduction potential of BzX measured via cyclic
voltametry (two cathodic peaks at −1.63 V and −1.91 V vs. Ag/AgCl), an exergonic electron transfer reac-
tion results. The ketyl radical resembles the well-known benzophenone ketyl. In conclusion, the triplet
3BzX* corresponds to an n → ␲* transition localized on the benzoyl substituent and resembles the ben-
zophenone triplet 3BP* but deviates from the triplet state of xanthone (3X*). This is supported through
DFT/B3LYP calculations, viz., (i) fully ground and triplet state geometry optimizations show that charge
and spin densities are localized on the benzoyl group, and (ii) calculation of the electronic transitions via
TD-DFT at B3LYP/631G+(d) and PB1BPE/631G+(d) levels of theory shows the local character. Agreement
with experiment is better by applying the conductor like polarized continuum model (CPCM) to consider
the solvent effect (MeCN).
Received 12 November 2010
Received in revised form 24 February 2011
Accepted 28 February 2011
Available online 8 March 2011
Keywords:
Photochemistry
Photopolymerization
Type II photoinitiator
Photoinitiator quenching
Electron transfer
The effectiveness of BzX as photoinitiator for the polymerization of MMA is found to be dou-
ble that of the unsubstituted xanthone (X). The photopolymerization rates (Rp) were found to be
7.06 × 10−4 mol L−1 s−1 in the case of BzX and 3.04 × 10−4 mol L−1 s−1 in the case of BX. This is attributed
to the fact that triplet 3BzX* behaves like the benzophenone triplet 3BP* and deviates from that of 3X*,
i.e., it is the localization of the triplet excitation on the benzoyl subunit which renders BzX a good type-II
photoinitiator.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
undergo direct photofragmentation upon absorption of light and
afford radicals capable of inducing polymerization (e.g. ben-
Photoinitiated polymerization is one of the most rapidly
expanding processes because of its widespread applications
in printing inks, adhesives, surface coating, microelectronics,
zoin ethers, acylphosphine oxides, hydroxyalkylphenyl ketones,
dialkoxy acetophenones, alkyl amino ketones, sulfonyl ketones,
silicon derivatives of ketones, etc.), and type II photoinitiating sys-
tems which are based on hydrogen transfer or electron transfer
followed by proton transfer. The latter process takes place between
the triplet excited state of the PI (usually an aromatic ketone, e.g.,
benzophenone and thioxanthone) and a coinitiator acting as hydro-
gen or electron/proton donor (e.g., a thiol or an amine), thereby
producing the initiating radicals. Some of the properties needed to
improve PI efficiency are high absorption coefficients, inefficient
quenching of the relevant excited states by the monomer or oxygen
and high quantum yields in producing the initiating radicals.
printing plates, etc. [1–9].
A crucial component in all pho-
topolymerization systems is the photoinitiator (PI) which absorbs
light and generates the active radicals to initiate the polymer-
ization. From the mechanistic point of view, there are two
main categories of photoinitiators. Type I photoinitiators, which
∗
Corresponding author. Tel.: +30 26510 08394; fax: +30 26510 08799.
E-mail address: msiskos@cc.uoi.gr (M.G. Siskos).
1010-6030/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2011.02.028

256
X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
described procedure [11a]. The preparations were carried out
11
10
O
12
13
under argon atmosphere using Schlenk techniques. Melting points
were taken on a Büchi 510 apparatus and are uncorrected. Infrared
spectra were recorded on a Perkin-Elmer FT-IR Spectrum GX as KBr
pellets. 1H and 13C NMR spectra were measured in CDCl3 solutions
on a Brucker AC-250 spectrometer. Chemical shifts (ı) are given
in ppm and referenced with respect to the residual signals of the
solvent, J values in Hz. Mass spectra were measured on an Agilent
1100 Series LC-MSD-Trap-SL spectrometer.
5
8
4
1
6
7
X: R = H
Ph
3
2
O
C
BzX: R =
9
R
O
Scheme 1.
Consequently, the role of the photoinitiator efficiency is very
2.2. Synthesis of 2-benzoyl-xanthone (BzX)
crucial and therefore the search for novel compounds that can
be used as photoinitiators has continuously received consider-
able attention in the last three decades. In order to improve
the photoinitiator properties some aspects of these studies are
focused in replacing the chromophore or in modifying the chro-
mophore through the introduction of suitable substituents (donors
or acceptors) on the same chromophore. A variety of substituted
benzophenones or thioxanthones have been examined for this pur-
pose and in some cases increased photoinitiation ability was found
[10].
Although many substituents in aromatic ketones have been
studied, nothing is known regarding the introduction of a second
carbonyl group. Thus, in this work we intend to extend the chro-
mophore of xanthone introducing a benzoyl group at the 2-position
(BzX, Scheme 1), to look for its potential function as photoinitia-
tor and to compare its activity against the unsubstituted xanthone.
More precisely, in this work we perform a systematic study which
focused: (i) to the photophysical and photochemical properties
using steady state absorption/emission and laser flash photolysis
techniques, (ii) to compare the photoinitiation activity of the substi-
tuted and unsubstituted xanthone (polymerization of MMA using
Et3N as co-initiator), (iii) to the quenching of the excited state of
PIs by triethylamine and monomer (MMA), and (iv) to shed light
to mechanistic questions related to the specific role played by each
carbonyl group. We have also performed quantum mechanical cal-
culations examining the geometry of the ground and triplet excited
states, as well as, the relevant molecular orbitals and electronic
transitions (PM3, AM1, Ab initio and DFT and TD-DFT) in order to
support the interpretation of the obtained results.
The present results show that 2-benzoyl xanthone (BzX) func-
tions as a type-II photoinitiator with an effectiveness twice as that
of the unsubstituted xanthone (X) for the polymerization of MMA.
As the spectral characteristics (photophysical, laser flash photoly-
sis and computational results) indicate, this is attributed to the fact
that triplet excitation of BzX is localized on the benzoyl subunit
rather than on the xanthonyl part. Thus, the triplet 3BzX* resem-
bles the benzophenone triplet 3BP* showing inefficient quenching
by the monomer MMA, undergoing, however, a very competitive
electron/proton transfer reaction with the co-initiator Et3N which
affords the desired initiating ␣-aminoethyl radical. In contrast, the
quenching of xanthone triplet 3X* by MMA is more competitive
compared to the crucial electron transfer step, diminishing thus its
polymerization ability.
A
hot (∼80 ◦C) solution of 0.5 g (1.75 mmol) of 2-
benzoylxanthene in 2.5 ml acetic acid was added slowly to a
stirred hot solution (∼80 ◦C) of 1.5 g (5.73 mmol) Na2Cr2O7 in
5 ml acetic acid. The mixture was heated at 100 ◦C for 30 min
until the oxidation reaction was completed (checked by TLC).
The resulting deeply green solution was left to cool and then
hydrolyzed. A greenish-yellow crude solid product was isolated
and recrystallized from a mixture of ethanol/chloroform (4:1)
giving pale yellow crystals (0.33 g, yield 64%); m.p.: 144–145 ◦C
(lit. [11b] m.p.: 146–147 ◦C); IR (KBr) (cm−1): 3062 (C–H, Ar), 1660
(C O, xanthone), 1651 (C O, ketone), 1605 (C C, Ar), 1464, 1345,
1316, 1277, 1265 (C–O), 1108, 757, 741; 1H NMR (CDCl3, 250 MHz,
298 K) ı (ppm): 8.70 (1H, d, J = 2 Hz, H1), 8.31 (1H, dd, J = 8.7, 2 Hz,
H3), 8.28 (1H, dd, J = 8.7, 2 Hz, H4), 7.78 (2H, dd, ortho-H), 7.73 (1H,
t, para-H), 7.70–7.30 (6H, m, H5,6,7,8 and meta-H); 13C NMR (CDCl3,
68.9 MHz, 298 K) ı (ppm): 194.9 (C O), 176.6 (C O), 158.5 (q),
156.0 (q), 137.2 (q), 135.7, 135.3, 133.2 (q), 132.7, 129.9, 128.9,
128.5, 126.8, 124.6, 121.8 (q), 120.9 (q), 118.8, 118.1; MS (APCI+)
m/z: 301.2 [M+H]+, MSMS (ESI+) m/z: 223 (M+-Ph) 104.7 [PhCO+].
2.3. Photopolymerization
Photopolymerizations were carried out in Pyrex tubes
(ꢀ > 300 nm) containing 3 ml of methylmethacrylate and the
proper amounts of initiator and triethylamine. They are filled with
dry argon prior to irradiation and are placed in a thermostatic bath.
An Osram 400 W high pressure lamp was used as light source.
At the end of the irradiation the reaction mixture was poured
into excess of cold methanol for precipitation. The polymer was
isolated by filtration and drying to constant weight in vacuum.
The monomer conversions for all samples were determined
gravimetrically.
2.4. UV absorption and fluorescence spectroscopy
Absorption spectra were recorded with a UV-vis Hitachi U-
2001 spectrophotometer. Fluorescence and excitation spectra were
obtained on a Cary Eclipse fluorescence spectrometer at 25 ◦C.
2.5. Steady-state UV photolyses
The photolyses were performed with a high pressure mercury
lamp (400 W) at 20 ◦C in quartz cuvettes. The photolysis solutions
(∼1 mM) were purged with argon or oxygen before irradiation.
2. Experimental
2.1. Materials
2.6. Electrochemistry
Acetonotrile, cyclohexane and MeOH (all spectroscopic grades)
were obtained from Merck and used without further purifica-
tion. Xanthone (99%, Aldrich) was recrystallized from ethanol,
the triethylamine (TEA, 99%, Aldrich) was distilled under vac-
uum. Methyl methacrylate (MMA, ≥99%, Fluka) was washed with
5% aq. NaOH solution, dried over CaCl2 and distilled in vacuum.
The 2-benzoylxanthene was prepared according to the previously
Cyclic voltammetric measurements were carried out using a
Metrohm instrument, Model 797 VA, Computerace equipped with a
three-electrode cell, containing a hanging mercury drop electrode
(HMDE) as a working electrode, a platinum wire as an auxiliary
electrode and an Ag/AgCl (3.0 mol L−1 KCl) reference electrode. The
measurements were performed under a nitrogen atmosphere at
25 ◦C in acetonitrile (ACN, Merck Uvasol reagent grade) containing

X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
257
0.1 M TBAPF6 (tetrabutylammonium hexafluorophosphate) as the
supporting electrolyte and at scan rate of 0.1 V s−1
.
2.7. ns-Laser flash photolysis
Solutions of compounds (A/cm ∼ 0.5) were deoxygenated by
bubbling with nitrogen or purged with oxygen and were photol-
ysed at 20 ◦C in a flow system (Suprasil quartz cell) by the fourth
harmonics (266 nm) of a Quanta-Ray GCR-11 Nd:YAG laser (Spec-
tra Physics, pulse duration <3 ns) with energies of 0.5–10 mJ. The
optical detection system consisted of a pulsed xenon lamp (XBO
150, Osram), a monochromator (Spectra Pro 275, Acton Research
Corporation), an R955 photomultiplier tube (Hamamatsu Photon-
ics) or a fast Si-photodiode with 1 GHz amplification, and a 1 GHz
digitizing oscilloscope (DSA 684A, Tektronix).
Scheme 2.
quencies and other important molecular properties of ground and
excited states.
Semi-empirical (PM3, AM1), ab initio and DFT calculations
were performed using the GAUSSIAN 98 or GAUSSIAN 03 program
[15]. The optimized geometry, vibrational modes, and vibrational
wavenumbers for the ground-state benzophenone, xanthone, BzX,
and their triplet states were obtained from DFT/B3LYP calculations
(restricted in the case of ground and unrestricted in the case of
triplet state and radicals) using the 6-31G(d) basis set. No imaginary
wavenumbers were observed at any of the optimized structures
shown here. The zero point vibrational energy and enthalpy correc-
tion were used without scaling. Time dependent-density functional
theory (TD-DFT) calculations were performed to estimate the elec-
tronic transition energies and their oscillator strengths.
2.8. Quantum yield of triplet formation (˚T)
Triplet quantum yield, ˚T, was determined by energy transfer
to perylene acting as triplet energy acceptor (ET = 147 kJ/mol) [12].
Optical matched MeCN solutions (deaerated) of BzX, xanthone (X)
and benzophenone (BP) as reference, were prepared in the pres-
ence of identical concentrations of perylene. The concentrations
should be high enough to guarantee quantitative quenching of the
triplets 3BzX*, 3X*, and 3BP* created using 266-nm laser light. The
quantum yield was calculated by comparing the T–T absorption
of the perylene triplet [13] produced by energy transfer from 3X*,
3BzX* and the reference (3BP*) according to the following equation
˚T(Y) = ˚T(3BP*) ꢁATPr(Y)/ꢁATPr(3BP*) ˚T(Y) is the triplet quan-
tum yield of 3BzX* and 3X*, while ˚T(3BP*) = 1 is the reference.
ꢁATPr(Y) and ꢁATPr(3BP*) are the corresponding absorbance of the
perylene triplet at 490 nm after triplet energy transfer from 3X*,
3BzX* and 3BP*, respectively.
3. Results and discussion
3.1. Synthesis
The synthesis of BzX was performed by a two step reaction pro-
cedure (Scheme 2). The first step is a Friedel-Crafts benzoylation
of xanthene using benzoyl chloride and AlCl3 as catalyst in CS2,
while the second step concerns a typical oxidation with Na2Cr2O7
in CH3COOH. Their structure were confirmed by 1H and 13C NMR,
FT-IR and MS spectra (see Section 2).
2.9. Measurement of the absorption coefficient (εT) of triplet
states
2.9.1. Actinometer method
This method is a comparative actinometry method using ben-
zophenone triplet (3BP*) in acetonitrile as reference [˚T(BP) = 1,
εT(BP) = 6500 L mol−1 cm−1 at 520 nm] [14]. Optically matched
samples of BP and BzX were irradiated with the 266 nm
laser light in nitrogen-saturated solutions. The triplet state
absorbance of BP at 520 nm and of target compound Y (BzX
3.2. Ground state geometry of BzX
Starting with low cost PM3 and AM1 semiempirical calculations
we found two different rotamers (BzXsyn and BzXanti), as expected
from the rotation of the benzoyl group. The first one has the two
carbonyl groups orientated in the same side of the molecule (syn)
and the second opposite (anti) (Fig. 1). Additionally, each rotamer
has an isoenergetic mirror image conformer BzX* (see Fig. S1, Sup-
plementary material). The optimized geometry of each conformer
found by semiempirical calculations (PM3 and AM1) was reopti-
mized by restricted HF and DFT method with the B3LYP function
and 6-31G and 6-31G(d) basis set. The optimized geometric param-
eters from the DFT calculation (bond lengths and angles) for BzX are
listed in Supplementary material, Table S1. From the two conform-
ers, the anti is predicted to be more stable by about 0.5–1.1 kcal/mol
depending on the calculation methodology used (see Table S2, sup-
plementary material). In all cases the xanthone moiety possesses a
planar geometry.
at 540 nm) were measured at different laser intensities.
A
plot of ꢁAT of BP and BzX as function of laser doses
a
[14b] was then used to calculate the absorption coefficient of
triplet–triplet absorption [εT(Y)] of compound according to equa-
tion εT(Y) = ˚T(BP)[εT(BP)ꢁAT(Y)]/[˚T(Y)ꢁAT(BP)], where εT is the
triplet molar absorption coefficient of BP or Y, and ꢁAT(BP) and
ꢁAT(Y) are the slopes of the triplet absorbance versus laser doses
for BP or Y, respectively.
2.9.2. Singlet depletion method
A second method is the singlet depletion technique which is
based on the well known relationship εT = εS ꢁODT/ꢁODS [14b]
where ꢁODS is the absorbance of the parent compound (negative
signal relative to the baseline) and ꢁODT that of the triplet state
appeared in the transient absorption spectrum. The εS was mea-
sured from the absorption spectrum. This method requires that no
triplet absorption overlaps with the ground state bleach.
3.3. UV–vis spectroscopy and TD-DFT calculations
The UV absorption spectra of BzX (Fig. 2) and xanthone X
(reference compound) were measured in the range of 10−3 to
10−6 (mol L−1) in MeCN, cyclohexane and MeOH and the values
of the molar absorption coefficients (ε) and the wavelength of
maximum absorptions (ꢀmax) are summarized in Table S3 (Supple-
mentary Information). The spectra of X and BzX are similar, except
of the ∼20 nm red-shift of the more intense band of X and BzX (from
2.10. Quantum mechanical calculations
Recent developments of computational chemistry constitute a
very useful tool for calculating geometries, energies, vibrational fre-

258
X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
Fig. 1. Optimized ground-state conformations of BzX calculated at the DFT/B3LYP/6-31G(d) level.
237 to 255 nm). As the solvent was changed from nonpolar cyclo-
hexane to polar MeCN or MeOH, all absorption bands were slightly
shifted toward longer wavelengths and were became broader. This
red-shift and the high values of ε is indicative of ␲ → ␲* type tran-
sitions. The carbonyl n–␲* transition should be expected to occur
on the tail of the absorption spectrum (for example in xanthone it
appears at ∼380 nm), but the low solubility of BzX in these solvents
prevented its observation.
Extensive spectroscopic as well as theoretical studies have been
devoted to xanthone [16–25]. Substitution on the 2-position (e.g.,
2-MeO-xanthone) does not change appreciably the basisity and the
␲-electron density on the carbonyl oxygen and the ipso-carbon
C2 [18]. We have applied the time dependent density functional
theory method (TD-DFT) for X and both conformers of BzX. This
method is able to predict absorption wavelengths corresponding
to vertical electronic transitions at a relatively small computer
time based on optimized ground state geometries and gives addi-
tionally an estimation of the solvent effect [26]. In our case, the
TD-DFT method was used with B3LYP and PBE1PBE functions and
6-31G+(d) basis set and the calculations were performed for vac-
uum/gas phase and in MeCN using the polarized continuum model
(CPCM) [27]. The results are presented and compared with experi-
mental data (Fig. 2, and for more detailed results see Supplementary
Information, Table S4). The B3LYP function gives better results for
the transitions in the area >300 nm, while the PBE1PBE function
reproduces the transitions well in <300 nm. In line with litera-
ture reports [18], we found very small p-orbital coefficients of the
ipso-carbon C2 regarding orbitals HOMO-3, HOMO-2, HOMO-1 and
LUMO involved in the transitions of X and BzX. As a consequence,
it is not surprising to find little spectral change by introducing the
2-benzoyl group in BzX and that the transitions are mostly local-
ized either in the xanthone or the benzoyl part. Characteristic is the
HOMO → LUMO transition which is localized on the xanthyl group
leaving the phenyl group almost uninvolved (Fig. 3).
While the xanthone molecule shown weak fluorescence spec-
trum, the emission spectra of BzX at room temperature were below
Fig. 2. Absorption spectra of BzX in acetonitrile (black line) and cyclohex-
ane (blue line). Electronic transition energies and the corresponding oscillator
strengths (gas phase) computed at TD-DFT/PBE1BPE/6-31G+(d) on the optimized
DFT/B3LYP/631G(d) ground state geometry (syn- and anti-rotamer). (For interpre-
tation of the references to colour in this figure legend, the reader is referred to the
web version of the article.)
Fig. 3. Calculated unoccupied (LUMO) and occupied (HOMO) molecular orbitals of
the anti rotamer of BzX at the B3LYP/6-31G+d level of theory.

X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
259
Fig. 4. (a) Time conversion curves for the bulk polymerization of MMA (9.42 mol L−1) at 30 ◦C using BzX or X as photoinitiator (5 × 10−3 mol L−1) and Et3N (0.1 mol L−1) as
co-initiator; (b) polymerization yield (%) as a function of Et3N concentration at 25 ◦C after 10 min irradiation using BzX as photoinitiator (5 × 10−3 mol L−1).
the detection limit of the conventional fluorescence spectrometer
used in the solvents like cyclohexane and MeCN.
3.4. Photopolymerization reaction
Polymerization was performed by irradiating a bulk solution of
BzX or X in MMA. Fig. 4a shows the time conversion curves of the
MMA polymerization using either X or BzX as photoinitiators with
Et3N as co-initiator at 30 ◦C. It is apparent that the polymerization
yield in both cases are increases linearly with the time, while the
polymerization rate (Rp), determined from the slopes of the straight
lines is more than two times greater in the case of BzX initiator
(Table 1). The dependence of the polymerization yield with tri-
ethylamine concentration after 10 min irradiation at 25 ◦C is shown
in Fig. 4b. The photoinitiation efficiency increases with the amine
concentration, reaching an optimum value at 0.02–0.1 M.
3.5. Nanosecond laser flash photolysis
3.5.1. Properties of the triplet state of BzX
Photoexcitation of a deoxygenated solution of BzX in acetoni-
trile with 266-nm (or 355-nm) laser light resulted in the formation
of a transient absorption signal with maxima at 325, 355, 530 nm
and a shoulder at ∼610 nm (Fig. 5, top) which we attribute to the
triplet state of BzX (3BzX*), see below. Additionally, a strong nega-
tive band (depletion of parent compound) appeared at wavelength
<300 nm with minimum at 260 nm (see Fig. 5 bottom). The decay of
transient signal can be fitted with first-order kinetics (see insets in
Fig. 5 top) at low excitation doses and has a lifetime of 22.7 ␮s. An
identical value was obtained also for the negative signal, a fact indi-
cating that the decay of the triplet leads to the regeneration of the
parent compound. The existence of an isosbestic point at 290 nm
supports further this suggestion. The linear dependence of the
absorption of this transient versus laser doses indicates that their
formation is a monophotonic process (Fig. 5 top, inset). At higher
excitation doses, the decay becomes faster and involves as expected
a second-order component as a result of triplet–triplet annihilation.
Fig. 5. Transient absorption spectra recorded after 266-nm laser excitation of BzX in
acetonitrile; in nitrogen saturated solution (top) and in oxygen saturated solution
(bottom). Inset (top): time profiles at 280 nm (recovery of ground state), at 315
Table 1
Photopolymerization of MMA with BzX in bulk at 30 ◦C.
and 530 nm and the photonity plot at 530 nm. Inset (bottom): decay at 530 nm and
quenching plot with oxygen.
[Initiator] (mol L−1
)
[TEA] (mol L−1
)
Rp (×104 mol L−1 s−1
)
X
BzX
0.1
0.1
3.04
7.06

260
X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
Fig. 8. Spin density distribution of triplet 3BzX* (anti rotamer).
3.5.2. Triplet state geometry and energy calculations
Fig. 6. Quenching of BzX triplet at 530 nm by 2-methylbutadiene and MMA in
As already shown (Section 3.2) BzX exhibits two different
rotamers (syn and anti). The triplet states of the two rotamers are
optimized at the unrestricted DFT/B3LYP/6-31G(d) level. Although
the optimized geometry of 3BzX* remains similar to the corre-
sponding ground state (see Fig. 7), there are substantial geometrical
changes located around the carbonyl group (C16–O17) of the ben-
zoyl substituent (see details in Table S1, Supplementary material).
The comparison reveals that the benzoyl carbonyl bond is length-
deaerated acetonitrile solution.
Under an oxygen atmosphere exactly the same transient was
recorded (Fig. 5 bottom), but its decay is accelerated following
pseudo first-order kinetics with a value of kobs = 2.2 × 107 s−1. By
taking into account the concentration of oxygen in MeCN at room
temperature ([O2] = 9.1 mM [12]), a bimolecular rate constant of
2.4 × 109 L mol−1 s−1 is evaluated, close to the quenching rate con-
stant of triplet xanthone (5.6 × 109 L mol−1 s−1 in benzene) [30a] or
triplet benzophenone with oxygen (2.3 × 109 L mol−1 s−1) [28], see
the inset in Fig. 5 bottom.
˚
ened by about 0.1 A, while the C2–C16 and C16–C18 bonds are
˚
shortened by about 0.056 and 0.051 A, respectively. Addition-
ally, the Mulliken atomic charges (Table 2) increase from −0.471
(ground state) to −0.330 (triplet state) on the oxygen (O17) and
decrease from 0.324 to 0.213 on the carbon atom (C16). On the
same time the spin density resides almost exclusively in the
benzoyl group, e.g. 0.938 (O17) versus −0.003 (O11), see Fig. 8
and Table 2. In contrast, geometry and charges of the xanthenyl
carbonyl C9–O11 group remain almost unchanged. The above
findings show the existence of an n→␲* triplet state (3BzX*)
localized on the benzoyl group in agreement with the laser exper-
iments and the localized character of the transitions described
above.
Further evidence characterizing this transient as triplet 3BzX*
is derived from its quenching with typical triplet quenchers like
2-methylbutadiene (isoprene) with ET = 60 kcal/mol [29], perylene
(ET = 35.1 kcal/mol [12]) and MMA. We found typical quench-
ing rate constants kq = 4.5 × 109 L mol−1 s−1 (2-methylbutadiene),
1.0 × 1010 L mol−1 s−1 (perylene) and 7.2 × 108 L mol−1 s−1 (MMA).
In Fig. 6 we show a representative quenching plot using the equa-
tion kobs = ko + kq[Q], where ko is the first order rate constant for
triplet decay in the absence of quencher Q.
Comparison of the absorption spectra of 3BzX* with that of xan-
thone triplet (3X*) shows that they are much different; 3X* has
been reported to show a maximum at about 630 nm in acetonitrile
[30a–k], see Table 3 for more details. On the contrary, the bands
observed for 3BzX* are almost similar to the corresponding spec-
trum of benzophenone triplet (3BP*) or its derivatives and this is a
first indication that the triplet excitation is localized mostly on the
benzoyl substituent and less to the xanthone moiety. This assump-
tion is further supported by the computational results reported
(Section 3.5.2).
We calculated with B3LYP/6-31G(d) the triplet energies of
3BzX*, 3X* and 3BP* using relaxed ground and triplet state
geometries and added ZPE corrections. We found 62.0 kcal/mol
for the anti and 61.1 kcal/mol for the syn conformer of 3BzX*,
68.3 kcal/mol for 3X* (experimental value 74.1 kcal/mol [28])
and 61.6 kcal/mol for 3BP* (experimental value 68.6 kcal/mol
[28]). The energy of 3BzX* is closer to benzophenone than to
xanthone, corroborating hitherto met conclusions for the local-
ization of triplet excitation in the benzoyl chromophores (see
Table 3).
Fig. 7. Optimized geometries of BzX (left) and 3BzX* (right) of the anti rotamer.

X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
261
Table 2
Mulliken atomic charges in ground state, triplet state and triplet state spin densities of BzX, calculated at the B3LYP/6-31G(d) level.
Atom
Ground state
Triplet state
Triplet spin density
Atom
Ground state
Triplet state
Triplet spin density
C1
C2
C3
C4
C5
C6
C7
C8
−0.239
0.071
−0.252
0.105
0.1893
−0.0447
0.1348
−0.0821
0.0039
0.0001
0.0056
0.0005
0.0067
C13
C14
C15
C16
O17
C18
C19
C20
C21
C22
C23
0.309
0.298
0.061
0.297
0.299
0.058
0.1458
−0.0021
0.0075
0.4426
0.9384
−0.150
−0.187
−0.182
−0.127
−0.134
−0.172
0.383
−0.552
−0.506
0.055
−0.170
−0.184
−0.183
−0.127
−0.135
−0.172
0.381
−0.551
−0.506
0.063
0.324
0.213
−0.471
−0.330
0.075
0.100
−0.0210
−0.173
−0.135
−0.118
−0.135
−0.150
−0.171
−0.135
−0.125
−0.132
−0.155
0.1374
−0.0605
0.1976
C9
O10
O11
C12
−0.0027
0.0134
−0.0619
−0.0934
0.1745
3.5.3. Determination of triplet state quantum yield and εT
3.5.4. The reaction of BzX triplet with Et3N and formation of
ketyl radical
Triplet state quantum yields were determined by using the
triplet energy transfer method to perylene. Perylene shows triplet
energy of 147 kJ/mol [13] and has a very low intersystem cross-
ing quantum yield (˚ISC = 0.011). Thus, perylene triplet can only
be formed through energy transfer from a suitable triplet donor.
Since perylene has a characteristic triplet–triplet absorption show-
ing a sharp peak at 490 nm, the quenching process could be easily
monitored either by following the triplet decay (3X* at 630 nm,
3BP* at 520 nm, and 3BzX* at 530 nm), or following the exponential
growth of the perylene triplet at 490 nm (see Fig. S4 in Supplemen-
tary material). Due to high triplet energy differences between the
donor and acceptor, a diffusion-controlled rate constant was found
for the triplet energy transfer of BzX: kq ∼ 1.0 × 1010 L mol−1 s−1
in MeCN, a fact which further confirms the triplet nature of the
transient obtained upon laser excitation of BzX (Fig. 5). Using BP
as standard (˚T = 1), a value of ˚T = 0.97 was found for xanthone
identical to the bibliographic value in CCl4 [30c]. For the BzX a
value of 0.78 and 0.81 was obtained using the 266 and 355 nm laser
excitation respectively, see Table 3.
Fig. 9 shows the transient absorption spectrum obtained after
266-nm laser excitation of the system BzX/Et3N in acetonitrile at
295 K (an analogous spectrum with 355-nm laser light was depicted
in Fig. S5 of Supplementary Material). The signal originally obtained
(80 ns after the pulse) is almost similar to the triplet state absorp-
tion (Fig. 4), and was later replaced (3 ␮s) by a different spectrum
showing maxima at 345 nm (with a shoulder at ∼330 nm) and
540 nm. The transient decays with first order-kinetics measured
at 620 nm (kobs = 1.8 × 106 s−1) and is followed by the build up of
the ketyl radical spectrum with maxima at 345 (kobs = 1.9 × 106 s−1
)
and 540 nm respectively (see the inset of Fig. 9). The ketyl radi-
cal is the well-known photoreduction product of carbonyl triplet
[42], while the expected a-aminoethyl radical (•CHMe-NEt2) is not
visible under our experimental conditions.
We produced the ketyl radical following a different path, by irra-
diating a solution of BzX in cyclohexane (Fig. 10). The spectrum
recorded immediately after the laser pulse is similar to the spec-
trum of the triplet state in MeCN and changes later (850 ns) to a
new transient which we attribute to the ketyl radical of BzX being
almost identical with that observed with Et3N in MeCN (3 ␮s, Fig. 9).
The decay of BzX ketyl radical at later time domain (second order)
and in oxygen atmosphere is presented at Supplementary Material
Fig. S6 and Fig. S7. The spectral characteristics of the BzX ketyl rad-
The triplet absorption coefficient of BzX at 530 nm was calcu-
lated with two different methodologies (see Section 2): the singlet
depletion methodology gave a value of 7900 L mol−1 cm−1 and
using benzophenone as actinometer a value of 7500 L mol−1 cm−1
was found.
Table 3
Spectral characteristics of the triplet states (absorption maxima in MeCN and cyclohexane, lifetimes, quantum yields, Tn ← T1 extinction coefficient and triplet energies) of
xanthone (X), benzophenone (BP) and BzX.
a
Compound
X
ꢀmax Tn ← T1 (nm)
ꢂtriplet (␮s)
˚
ε of Tn ← T1 (L mol−1 cm−1
)
ET (kcal/mol)
T
MeCN
CH
MeCN
CH
MECN
635
620
630
631
627
520
315
355
530
610
0.02
0.90d
0.97f
5300b
8600c
9500d
74.1g
73d
16.8
8.3
0.95
14 ␮s
22.7
72o
0.022
BP
BzX
535
315
340
530
0.3n
0.105
1
6500e
7500l
7900m
68.6 and 69.2i
68.5o
0.78j
0.81k
a
Ref. [30a–k].
In benzene Ref. [30a].
In benzene Ref. [34].
In EtOAc Ref. [35].
In MeCN Ref. [14].
In CCl4 Ref. [30e].
Refs. [17,36,37].
Ref. [12], p.107.
b
c
d
e
f
g
i
j
With 266 nm excitation laser light.
With 355 nm excitation laser light.
BP actinometry.
k
l
m
n
o
Singlet depletion.
Ref. [41].
T1 ← So from TDDFT calculations (see text).

262
X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
The above characteristics reflect the n–␲* character of BzX
triplet, in agreement with the TD-DFT calculations for the So→T1
transition which takes place at 415.6 nm in the gas phase and at
401.7 nm in MeCN; this blue shift supports further the n–␲* char-
acter of the transition.
The bimolecular quenching rate constant of 3BzX*
with Et3N (electron transfer) in MeCN was measured,
kq = 3.5 × 109 L mol−1 s−1 (Fig. S9 of Supplementary Material)
and was found to be very close to corresponding literature values
of 3BP* (kq = 3.5 × 109 L mol−1 s−1 [39] and 4.0 × 109 L mol−1 s−1
[40] respectively). A slightly greater value was reported in the
case of 3X* in the same solvent (kq = 4.4 × 109 L mol−1 s−1 [39] and
2.4 × 109 L mol−1 s−1 [30k]). The observed values are in accordance
to the free energy values (ꢀGet) which are obtained applying the
Rehm–Weller equation [31] and using the measured redox poten-
tials of the X and BzX and literature values of X and BP [32,33]. A
free energy value (ꢁGet) of −10.6 kcal/mol was obtained for the
electron transfer from NEt3 to 3BzX* and a value of −13.8 kcal/mol
for the 3X* (see the electrochemical details in Supplementary
Material).
Fig. 9. Transient absorption spectra observed in the laser photolysis of BzX
(1.35 × 10−5 mol L−1) and Et3N (2.9 × 10−4 mol L−1) with 266 nm laser light in nitro-
gen saturated MeCN; (ꢀ) 80 ns, (ꢁ) 300 ns, (ꢂ)1.50 ␮s and (♦) 3.0 ␮s after the laser
pulse. Inset: (a) the build up of ketyl radical at 350 nm and (b) normalized decay
profiles at 540 (black line), at 620 nm (blue line) and the simulated build up of ketyl
radical at 540 nm (green line). (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of the article.)
3.5.5. Conclusions
The present investigation shows that introduction of the ben-
zoyl group in xanthone (X) produces the 2-benzoyl xanthone (BzX)
which enhances the photoinitiator activity of methylmethacrylate
(MMA) using Et3N as co-initiator (Table 1). The spectroscopic and
theoretical results presented above, indicate a type-II photoinitia-
tion system which follows the typical mechanistic scheme for the
photoreduction of aromatic ketones by amines, as has been estab-
lished by numerous studies [42]. Thus, after excitation in polar
solvents like acetonitrile, a contact ion pair (CIP) was formed via
electron transfer between the triplet 3BzX* and Et3N (Scheme 4, a).
This triplet contact ion pair follows three possible decay pathways:
to undergo proton transfer generating a triplet radical pair (ketyl
and ␣-aminoethyl radical, pathway b), to form a solvent separated
ion pair (SSIP, step c), or to undergo back-electron transfer regen-
erating the reactants (step d). The resulting ␣-aminoethyl radical
initiates finally the polymerization of the monomer MMA.
Upon excitation of BzX with the 266 nm or 355 nm laser
light in MeCN solvent, the triplet state 3BzX* was produced and
the quantum yield (˚ = 0.80) and its molar absorption coeffi-
cient was determined. Based on, (i) the spectral characteristics of
the transient spectrum (Fig. 5, Section 3.5.1), (ii) the quenching
rate constants measured with typical triplet quenchers (Section
3.5.3, Fig. 6), (iii) the observed photoreduction with Et3N (Sec-
tion 3.5.4) and the reduction potentials determined hereupon
(Supplementary Material), as well as (iv) the TD-DFT calculations
(Section 3.5.2), it was concluded that 3BzX* corresponds to an
n→␲* transition localized almost exclusively on the benzoyl sub-
stituent; the close similarity to the benzophenone triplet 3BP* and
the disparity from 3X* strengthens this conclusion further. In other
words, the two subunits xanthonyl and benzoyl seem to function as
localized chromophores, and, irrespective of which chromophore
is originally excited, the lower energy triplet state is finally pop-
ulated through intramolecular energy transfer. Latter is facilitated
through the ca. 6 kcal/mol reduction of the triplet energy of 3BzX*
(being similar to 3BP*) compared to 3X* (Section 3.5.2).
Fig. 10. Transient absorption spectra recorded after 266 nm laser excitation of BzX
in cyclohexane solution at (᭹) 60 ns, (ꢀ) 200 ns and (ꢂ) 850 ns after the laser pulse.
Inset: decay at 600 (blue line) and 540 nm (red line) and the build up of ketyl radical
at 350 nm (black line) with normalized amplitude. (For interpretation of the refer-
ences to colour in this figure legend, the reader is referred to the web version of the
article.)
ical are similar to the benzophenone ketyl radical but differ from
the xanthone ketyl; the latter has a maximum at about 450 nm and
a broad absorption which extends until 800 nm [30a,e,g,k] which
is absent in BzX ketyl. Supportive evidence comes from open-shell
DFT/B3LYP/6-31G(d) calculations which showed that the benzoyl
ketyl type radical (I) is more stable than the xanthyl ketyl radical
(II) by about 3 kcal/mol (Scheme 3).
Obviously, it is just this triplet character and the localization on
the benzoyl group which doubles the photoinitiation activity of BzX
compared to X (Table 1) and despite the more favourable driving
force (more exergonic ꢁGet) of the primary electron transfer event
(step a, Scheme 4) for the latter (X). The reason seems to be the
fact that the monomer (MMA) functions as a quencher for triplet
states: it is much more effective for 3X* (kq = 1.2 × 1010 L mol−1 s−1
O
O
O
Ph
Ph
C
C
O
OH
OH
I
II
[10b]) than for 3BzX* (kq = 7.2 × 108 L mol−1 s−1
, see Section
Scheme 3.
3.5.1). For comparison, the quenching of benzophenone triplet

X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
263
kisc
1BzX* + NEt3
3BzX* + NEt3
ket
kdiff
3
a
BzX
NEt3
BzX
+
NEt3
c
contact ion pair
solvent separated ion pairs
hv
kbet
kpt
b
d
X
BzX + NEt3
Me
polymerization
+
OH
C
NEt2
H
Ph
α-amino ethyl
ketyl
radical
radical
Scheme 4.
3BP* by MMA (kq = 5.4 × 107 L mol−1 s−1 in MeCN [10b] and
kq = 6.9 × 107 L mol−1 s−1 in benzene [38]) was found to be close
to that of 3BzX*. In contrast, the quenching with Et3N (the
primary electron transfer step a), does not differentiate much
for 3X* (kq = 4.4 × 109 L mol−1 s−1 [39] and 2.4 × 109 L mol−1 s−1
[11] (a) V. Georgakilas, G.P. Perdikomatis, A.S. Triantafyllou, M.G. Siskos, A.K.
Zarkadis, Tetrahedron 58 (2002) 2441–2447;
(b) J. Heller, St. v. Kostanecki, Ber. Dtsch. Chem. Ges. 41 (1908) 1324–1327.
[12] S.L. Murov, I. Carmichael, G.L. Hug, Carmichael, Handbook of Photochemistry,
2nd ed., Marcel Dekker, New York, 1993, p. 57.
[13] (a) S.A. Yamamoto, K. Kikuchi, H. Kokubun, J. Photochem. 7 (1977) 177–184;
(b) R.V. Bensasson, E.J. Land, Trans. Faraday Soc. 67 (1971) 1904–1915.
[14] (a) R.V. Bensasson, J.C. Gramain, J. Chem. Soc., Faraday Trans. 176 (1980)
1801–1810;
(b) I. Carmichael, G.L. Hug, J. Phys. Chem. Ref. Data 15 (1986) 1–250.
[15] (a) M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheese-
man, V.G. Zakrzewski, J.A. Montgomery Jr., R.E. Stratmann, J.C. Burant, S.
Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford,
J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, B.B. Stefanov, G.
Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, M. Challa-
combe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C. Gonzalez,
M. Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian 98 Revision A.9, Gaussian,
Inc, Pittsburgh, PA, 1998;
[30k] in MeCN), for 3BzX* (kq = 3.5 × 109 L mol−1 s−1
, Section
3.5.4 and Fig. S9 of Supplementary Material) or for 3BP*
(kq = 3.5 × 109 L mol−1 s−1 [39], 4.0 × 109 L mol−1 s−1 [40] in MeCN).
Thus, the quenching of xanthone triplet by MMA is a very com-
petitive process compared to the crucial electron/proton transfer
reaction which affords the desired ␣-aminoethyl radical through
steps a and b. On the contrary, the electron transfer is more com-
petitive in the cases of 3BzX*and 3BP*.
Summarizing, the above results show that triplet 3BzX* behaves
like the benzophenone triplet 3BP* and deviates from that of xan-
thone 3X*, i.e., triplet excitation is localized mostly on the benzoyl
subunit of BzX. This property renders 2-benzoyl-xanthone a much
more effective type-II photoinitiator than the unsubstituted xan-
thone.
(b) M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheese-
man, J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P.
Hratchian, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Strat-
mann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K.
Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich,
A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari,
J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Ste-
fanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith,
M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. John-
son, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03 Revision D.01,
Gaussian, Inc., Wallingford, CT, 2004.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jphotochem.2011.02.028.
[16] F. Scigalski, J. Paczkowski, Macromol. Chem. Phys. 209 (2008) 1872–1880.
[17] (a) H.J. Pownall, J.R. Huber, J. Am. Chem. Soc. 93 (1971) 6429–6436;
(b) H.J. Pownall, R.E. Connor, J.R. Huber, Chem. Phys. Lett. 22 (1973) 403–405.
[18] K. Mizutani, K. Miyazaki, K. Ishigaki, H. Hosoya, Bull. Chem. Soc. Jpn. 47 (1974)
1596–1603.
[19] A.A. Efimov, R.N. Nurmukhametov, I.L. Belaits, A.I. Tolmachev, Opt. Spectrosc.
30 (1971) 337.
[20] T. Minegishi, T. Hoshi, H. Hiratsuka, Y. Tanizaki, Bull. Chem. Soc. Jpn. 50 (1977)
3140–3143.
[21] R.E. Connors, R.J. Sweeney, F. Cerio, J. Phys. Chem. 91 (1987) 819–822.
[22] S. Ishijima, M. Higashi, H. Yamaguchi, J. Phys. Chem. 98 (1994) 10432–10435.
[23] S.L. Zheng, P. Coppens, Chem. Eur. J. 11 (2005) 3583–3590.
[24] A. Thöny, M.J. Rossi, J. Photochem. Photobiol. A: Chem. 104 (1997) 25–33.
[25] B. Heinz, B. Schmidt, C. Root, H. Satzger, F. Milota, B. Fierz, T. Kiefhaber, W.
Zintha, P. Gilch, Phys. Chem. Chem. Phys. 8 (2006) 3432–3439.
References
[1] X. Asvos, PhD thesis in preparation, University of Ioannina, Greece.
[2] (a) J.P. Fouassier, J.F. Rabek (Eds.), Radiation Curing in Polymer Science and
Technology, vols. I–IV, Elsevier Applied Science, London, 1993;
(b) J.P. Fouassier, Photoinitiation, Photopolymerization and Photocuring Fun-
damentals and Applications, Hanser Publishers, New York, 1995 (Chapter 1).
[3] C.N. Bowman, C.J. Kloxin, AIChE J. 58 (2008) 2775–2795.
[4] H.F. Gruber, Prog. Polym. Sci. 17 (1992) 953–1044.
[5] A.B. Scranton, C.N. Bowman, R.W. Peiffer (Eds.), Photopolymerization Fun-
damentals and Applications, ACS, Washington, 1997, ACS Symposium Series
673K.
[6] W. Rutsch, K. Dietliker, D. Leppard, M. Köhler, L. Misev, U. Kolczak, G. Rist, Prog.
Org. Coat. 27 (1996) 227–239.
[26] (a) M. Cossi, V. Barone, J. Chem. Phys. 115 (2001) 4708–4717;
(b) D. Jacquemin, J. Preat, E.A. Perpete, Chem. Phys. Lett. 410 (2005) 254–259;
(c) D. Jacquemin, J. Preat, M. Charlot, V. Wathelet, J.M. Andre, E.A. Perpete, J.
Chem. Phys. 121 (2004) 1736–1743.
[27] (a) S. Miertus, E. Scrocco, J. Tomasi, Chem. Phys. 55 (1981) 117–129;
(b) S. Miertus, J. Tomasi, Chem. Phys. 65 (1982) 239–241;
(c) B. Mennucci, E. Cances, J. Tomasi, J. Phys. Chem. 101 (1997) 10506–10517;
(d) M. Cossi, V. Barone, J. Cammi, Chem. Phys. Lett. 255 (1996) 327–335.
[28] S.K. Chattopadhyay, C.V. Kumar, P.K. Das, J. Photochem. 30 (1985) 81–91.
[29] R.E. Kellogg, W.T. Simpson, J. Am. Chem. Soc. 87 (1965) 4230–4234.
[7] J.P. Fouassier, X. Allonas, D. Burget, Prog. Org. Coat. 47 (2003) 6–36.
[8] R.S. Davidson, A.A. Dias, D. Illsley, J. Photochem. Photobiol. A: Chem. 89 (1995)
75–87.
[9] D.A. Tasis, M.G. Siskos, A.K. Zarkadis, Macromol. Chem. Phys. 199 (1998)
1981–1987.
[10] (a) N.S. Allen, N.G. Salleh, M. Edge, T. Corrales, M. Shah, F. Catalina, W.A. Green,
Eur. Polym. J. 33 (1997) 1639–1643;
(b) H.J. Timpe, K.P. Kronfeld, J. Photochem. Photobiol. A: Chem. 46 (1989)
253–267;
(c) H.J. Timpe, K.P. Kronfeld, U. Lammel, J.P. Fouassier, D.J. Lougnot, J. Pho-
tochem. Photobiol. A: Chem. 52 (1990) 111–122;
(d) L. Cokbaglan, N. Arsu, Y. Yagci, S. Jockusch, N.J. Turro, Macromolecules 36
(2003) 2649–2653;
(e) G. Temel, N. Arsu, J. Photochem. Photobiol. A: Chem. 202 (2009) 63–66;
(f) M. Aydin, N. Arsu, Y. Yagci, Macromol. Rapid Commun. 24 (2003) 718–723;
(g) M. Aydin, N. Arsu, Y. Yagci, S. Jockusch, N.J. Turro, Macromolecules 38 (2005)
4133–4138.
[30] (a) A. Garner, F. Wilkinson, J. Chem. Soc., Faraday Trans.
1010–1020;
2 (72) (1976)
(b) F. Wilkinson, A. Garner, J. Chem. Soc., Faraday Trans. 2 (73) (1977) 222–223;
(c) F. Wilkinson, A. Garner, Photochem. Photobiol. 27 (1978) 659–670;
(d) C. Ley, F. Morlet-Savary, J.P. Fouassier, P. Jacques, J. Photochem. Photobiol.
A: Chem. 137 (2000) 87–92;
(e) J.C. Scaiano, J. Am. Chem. Soc. 102 (1980) 7747–7753;

264
X. Asvos et al. / Journal of Photochemistry and Photobiology A: Chemistry 219 (2011) 255–264
(f) C.H. Evans, N. Prud’Homme, M. King, J.C. Scaiano, J. Photochem. Photobiol.
(b) J.B. Guttenplan, S.B. Cohen, J. Am. Chem. Soc. 94 (1972) 4040–4042.
A: Chem. 121 (1999) 105–110;
[34] T. Werner, J. Phys. Chem. 83 (1979) 320–325.
(g) C. Coenjarts, J.C. Scaiano, J. Am. Chem. Soc. 122 (2000) 3635–3641;
(h) H. Satzger, B. Schmidt, C. Root, W. Zinth, B. Fierz, F. Krieger, T. Kiefhaber, P.
Gilch, J. Phys. Chem. A 108 (2004) 10072–10079;
(j) K.A. Abdullah, T.J. Kemp, J. Photochem. 32 (1986) 49–57;
(k) J.-T. Wang, Y. Pan, L.-M. Zhang, S.-Q. Yu, Chin. J. Chem. Phys. 20 (2007)
395–400.
[35] P.B. Merkel, J.P. Dinnocenzo, J. Photochem. Photobiol. A: Chem. 193 (2008)
110–121.
[36] (a) R.E. Connors, W.R. Christian, J. Phys. Chem. 86 (1982) 1524–1528;
(b) M. Baba, T. Kamei, M. Kiritani, S. Yamauchi, N. Hirota, Chem. Phys. Lett. 185
(1991) 354–358;
(c) Y. Ohshima, T. Fujii, T. Fujita, D. Inaba, M. Baba, J. Phys. Chem. A 107 (2003)
8851–8855.
[31] (a) D. Rehm, A. Weller, Ber. Bunsen-Ges. Phys. Chem. 73 (1969) 834–
839;
(b) D. Rehm, A. Weller, Isr. J. Chem. 8 (1970) 259–271.
[32] (a) M. Lucarini, G.F. Pedulli, A. Alberti, C. Paradisi, S. Roffia, J. Chem. Soc., Perkin
Trans. 2 (1993) 283–287;
[37] K. Bartl, A. Funk, M. Gerhards, Chem. Phys.Chem. 10 (2009) 1882–1886.
[38] R. Kuhlmann, W. Schnabel, Angew. Makromol. Chem. 57 (1977) 195–210.
[39] N. Hoffmann, H. Görner, Chem. Phys. Lett. 383 (2004) 451–455.
[40] J.M. Figuera, R. Sastre, A. Costela, I. Garcia-Moreno, M.T. Al-Hakakk, J. Dabrio,
Laser Chem. 15 (1994) 33–46.
[41] G. Porter, M.R. Topp, Proc. R. Soc. Lond., Ser. A 315 (1970) 163–184.
[42] See Refs. [1–38] in Supplementary material which are related to photoreduc-
tion of aromatic ketones.
(b) A. Streitwieser Jr., I. Schwager, J. Phys. Chem. 66 (1962) 2316–2320;
(c) S. Bonafede, M. Ciano, F. Bolletta, V. Balzani, L. Chassot, A. Von Zelewsky, J.
Phys. Chem. 90 (1986) 3836–3841.
[33] (a) P.J. Wagner, R.J. Truman, A.E. Puchalski, R. Wake, J. Am. Chem. Soc. 108
(1986) 7727–7738;