European Polymer Journal 136 (2020) 109909

Contents lists available at ScienceDirect

European Polymer Journal

journal homepage: www.elsevier.com/locate/europolj

Photophysical, kinetic and thermodynamic study of one-component Type II T

thioxanthone acetic acid photoinitiators
Eyup Metina,c, Nergis Arsub, Saron Cataka, , Viktorya Aviyentea,
⁎ ⁎

a
Bogazici University, Department of Chemistry, 34342 Bebek, Istanbul, Turkey
b
Yildiz Technical University, Department of Chemistry, 34220 Davutpasa, Istanbul, Turkey
c
Turkish-German University, Department of Materials Science and Technology, 34820 Beykoz, Istanbul, Turkey

ARTICLE INFO ABSTRACT

Keywords: In this study, One-Component Type II thioxanthone acetic acid based photoinitiators, as well as the conventional
Thioxanthone Type II photoinitiator with phenyl acetic acid based co-initiators, have been investigated in terms of their
DFT photophysical, kinetic and thermodynamic properties by a series of Density Functional Theory (DFT) calcula­
Photoinitiator tions. Experimental UV–Vis spectra have been compared with computationally generated UV–Vis spectra via
Fukui function
Time-Dependent DFT (TD-DFT) and by taking into consideration the vibrational and dynamic effects (Wigner).
Spin density
Natural transition orbitals
Natural transition orbitals have been generated for thioxanthone acetic acid derivatives to understand the
physical nature of electronic transitions. Furthermore, kinetic and thermodynamic parameters of radical attack
to the monomer of thioxanthone acetic acid based radicals and phenyl acetic acid based co-initiator radicals have
been investigated. Comparison of the thermodynamic parameters of the initiating radicals has shown that in­
itiation is more favorable with exocyclic radicals in both cases. Exocyclic radicals of the thioxanthone acetic acid
derivatives and phenyl acetic acid derivates have been found to have comparable activation energies for their
attack to methyl methacrylate. Furthermore, the spin densities and the radical Fukui functions of the attacking
radicals have been nicely correlated with the activation barriers. In this way, a fast and predictive approach for
the determination of radical reactivity in photopolymerization reactions has been proposed; this enables the
evaluation of the initiation efficiency of the photoinitiators prior to synthesis.

1. Introduction initiator to form radicals and the radicals that are responsible for the
initiation of the polymerization are formed on the co-initiator mole­
Photopolymerization is closely related to polymer science and cules. On the other hand, there is another type of photoinitiator that can
technology due to its wide variety of applications such as UV-curing, be classified under Type II classification as One-Component Type II
coatings, adhesives, inks, dental materials, 3D-printing, biomedical and photoinitiator. This type of photoinitiator does not require a co-initiator
biochemical applications [1–7]. Photoinitiators can be monomeric or and is capable of making intermolecular or intramolecular hydrogen
polymeric [2,5]. Upon irradiation with light, the following reactions abstraction; radical species responsible for the initiation of the poly­
may occur; merization are formed on the photoinitiator and the reaction proceeds
from these radicals [5–8].
(i) fragmentation by photo cleavage of the bonds, The most commonly used Type II photoinitiators are camphorqui­
(ii) transfer of the energy to another molecule to form radicals that none, benzophenone and thioxanthone (Fig. 1) and they need a co-in­
initiate the polymerization, itiator to initiate the photopolymerization [4,7–9]. Among them
(iii) abstraction of hydrogen from another molecule and thus creating a thioxanthone and its derivatives are widely used as photoinitiators for
radical which will initiate the reaction. photoinitiated polymerization reactions, their efficiencies and initiating
mechanisms are deeply investigated by experimental methods [10–18].
Photoinitiators that form radicals by photo-fragmentation are called Mercapto functionalized thioxanthone is an early example of One-
Type I and those that form radicals by hydrogen abstraction are called Component Type II photoinitiators [19]. Beside the photoinitiating
Type II photoinitiators. Type II photoinitiators need a synergist / co- capabilities of thioxanthone and thioxanthone based One-Component

⁎
Corresponding authors.
E-mail addresses: [email protected] (S. Catak), [email protected] (V. Aviyente).

https://doi.org/10.1016/j.eurpolymj.2020.109909
Received 30 June 2020; Accepted 20 July 2020
Available online 25 July 2020
0014-3057/ © 2020 Elsevier Ltd. All rights reserved.
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Fig. 1. (a) Thioxanthone (TX), (b) Thioxanthone acetic acid (TX-AA), (c) Thioxanthone acetic acid derivatives (TX-NH-AA, TX-O-AA, TX-S-AA), (d) Phenyl Acetic
acid (PHC), (e) Phenyl acetic acid derivatives (PHN, PHO, PHS), (X: -NH, -O, -S).

Type II photoinitiators [15–20], the former are also used in a variety of Phenyl acetic acid derivatives, especially N-phenylglycine, are well-
studies such as the photochemical preparation of metal nanoparticles known co-initiators which are used in Type II photoinitiation systems
[3,21], DNA binding mode studies [22], functionalization of polyhedral [28–30]. Non-functionalized thioxanthone (Fig. 1) is capable of Type II
oligomeric silsesquioxanes (POSS) [23], nanocomposite preparations photoinitiation [8]. The reaction mechanism of non-functionalized
[21,24], and recently as candidates for thermally activated delayed thioxanthone molecules with phenyl acetic acid based co-initiators is
fluorescence (TADF) emitter designs [25–27]. displayed in Fig. 4. In contrast to One-Component Type II initiation
The capability of light absorption of a molecule plays a significant mechanisms of thioxanthone acetic acid derivatives, the conventional
role in absorbing and utilizing the energy for radical formation. Type II photoinitiation reactions of phenyl acetic acid derivatives with
Thioxanthone acetic acid derivatives have π electrons thus, their high thioxanthone lead to the formation of endocyclic radical on the thiox­
molar absorption ability in the UV–Vis region can be attributed to the anthone core (red) and an exocyclic radical on the co-initiator (blue)
delocalization of electrons (Fig. 1). Recent studies on photoinitiators (Fig. 4).
focus on introducing new functionalities to both polymeric and In photopolymerization reactions acrylates are one of the mostly
monomeric photoinitiators to enhance their absorptivity in the visible used monomers because the radicals are capable of attacking the C]C
region as well as their photo-reactivity [5,7]. Hence, understanding the double bond and initiating the free radical polymerization [5–8,15–20].
physical, chemical, photophysical, kinetic and thermodynamic prop­ Thioxanthone based photoinitiators and one of the widely known
erties of photoinitiators has a crucial impact on providing detailed in­ industrial Type I photoinitiators, Irgacure 907, have been investigated
formation on novel photoinitiators. with computational tools [31–37]. Even though there is an increased
Experimentally proposed radical formation reaction mechanisms of interest on photopolymerization reactions and thioxanthone deriva­
thioxanthone acetic acid derivatives are displayed in Fig. 2 and Fig. 3. tives, to the best of our knowledge, the computational studies on these
In both reactions, upon irradiation by a light source, thioxanthone species are limited.
acetic acid derivatives are known to reach a triplet state and make an In this study, the initiation capabilities of four different thiox­
intermolecular hydrogen abstraction spontaneously which further re­ anthone based One-Component Type II Photoinitiators; 2-(9-oxo-9H-
sults in decarboxylation to evolve carbon dioxide yielding two radicals: thioxanthen-2-yl)acetic acid (TX-AA) [15], (9-oxo-9H-thioxanthen-2-
(i) exocyclic (blue) and (ii) endocyclic (red). yl)glycine (TX-NH-AA) [16], 2-((9-oxo-9H-thioxanthen-2-yl)oxy)acetic
Arsu et al. have deeply investigated the photopolymerization of acid (TX-O-AA) [17,18] and 2-((9-oxo-9H-thioxanthen-2-yl)thio)acetic
these species with experimental tools such as laser flash photolysis, and acid (TX-S-AA) [17,18] have been investigated. Furthermore, phenyl
have proposed that photoinitiation takes place preferentially via the acetic acid derivatives; 2-phenylacetic acid (PHC), N-phenylglycine
exocyclic radical (blue) whereas the endocyclic radical (red) leads to (PHN), 2-phenoxyacetic acid (PHO) and 2-(phenylthio)acetic acid
side products [15–17]. (PHS) have also been studied to compare the advantages of One-

Fig. 2. Experimentally proposed radical formation reaction of thioxanthone acetic acid (TX-AA) [15].

2
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Fig. 3. Experimentally proposed radical formation reaction of thioxanthone acetic acid derivatives (TX-X-AA, X: -NH, -O, -S) [16,17].

Component Type II photoinitiators vs. Type II photoinitiators (Fig. 1). To levels of theory have been carried out. 6-31+G(d,p) basis set has been
reach this goal the photophysical, kinetic and thermodynamic proper­ used for all atoms except S, for which the 6-311++G(3df,3pd) basis set
ties of these species have been calculated, the UV–Vis spectra and has been added. Unrestricted formalism has been utilized for all open-
natural transition orbitals are generated. Then, the kinetic and ther­ shell species during calculations. The stationary points for all minima
modynamic parameters of the radical formation and radical attack to have been characterized with positive frequencies, whereas transition
methyl methacrylate reactions have been modeled. Finally, spin density structures have an imaginary frequency.
and radical Fukui functions of the radicalic species have been correlated The electronic absorption spectra of thioxanthone acetic acid deri­
with the activation barriers with the aim of evaluating the initiation vatives have been calculated on the optimized geometries with the
efficiency of the photoinitiators prior to synthesis. Time-Dependent Density Functional Theory (TD-DFT) at the same le­
vels of theory.
Frontier orbitals were generated via TD-DFT calculations with the
2. Methodology Gaussian 09 (Rev. E.01.) software package and visualized with the
Avogadro software package [45].
2.1. Density functional theory In order to include dynamic and vibrational effects, 20 snapshots
have been obtained by calculating the harmonic vibrational frequencies
All calculations have been carried out with the Gaussian 09 (Rev. via Wigner distribution as implemented in the Newton-X Ver. 2 soft­
E.01.) software package [38]. Density Functional Theory (DFT) calcu­ ware package [46] and these have been used to generate UV–Vis
lations at the B3LYP [39–41], BMK [42], M06-2X [43], ωB97X-D [44]

Fig. 4. Radical formation reaction of phenyl acetic acid (PHC) and phenyl acetic acid derivatives (PHX, X: -NH, -O, -S).

3
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Fig. 5. Optimized geometries of (a) TX-AA, (b) TX-NH-AA, (c) TX-O-AA and (d) TX-S-AA) (B3LYP/6-31+G(d,p) and 6-311++G(3df,3pd) for S atom in DMF).

spectra for all thioxanthone acetic acid derivatives. Equation [55]:

kB T RT G‡/ RT
2.2. Solvent model k= · ·e
h Po (4)
−1 −1
Since the experimental studies for thioxanthone acetic acid deriva­ In Equation (4), k is reaction rate constant (L.mol .s ), kB is
tives have been conducted in N,N-dimethylformamide (DMF), the in­ Boltzmann’s constant (1.38065 × 10−23 J.K−1), T is temperature
tegral equation formalism polarizable continuum model (IEF-PCM) has (298.15 K), h is Planck’s constant (6.62608 × 10−34 J.s), R is ideal gas
been used to mimic the solvent environment. DMF (ε = 37.219) has constant (8.31447 J.K−1.mol−1 and 8.20574 × 10−2 L.atm.
been used as the solvent both in the DFT and TD-DFT calculations for all K−1.mol−1), Po is the pressure (1 atm) and ΔG‡ is the Gibbs Free Energy
species. of activation [55].

2.3. Natural transition orbitals (NTO) 2.6. Geometry optimizations

Natural Transition Orbitals (NTOs) and Φs values have been eval­ Optimizations have been carried out with B3LYP/6-31+G(d,p) and
uated with the Nancy_EX tool [47–49] and the generated NTOs have 6-311++G(3df,3pd) for S atom in DMF. After preparing the initial
been visualized with the Avogadro software package [45]. input of the structure, preliminary optimizations have been carried out:
360° dihedral scans (12 steps with 30°) have been conducted for all
2.4. Natural Population Analysis (NPA) and radical Fukui functions possible rotations around single bonds. Subsequent to scans, the
minimum energy conformers have been selected for further optimiza­
Electron density of atomic sites have been calculated with Natural tions to obtain the geometries corresponding to stationary points. In
Population Analysis (NPA). The optimized ground state geometries of conjunction with the transition state search, intrinsic reaction co­
the neutral species have been used for the single-point calculations of ordinate (IRC) calculations have been conducted for the verification of
the cationic and the anionic species. the transition state structures. Optimization of the conformers gener­
Fukui functions on a kth atomic site of a molecule can be calculated ated by IRC calculations have yielded, the pre-reactive complex and the
by the following equations for nucleophilic attack ( f k+) (Eq. (1)), for intermediate structures. All image renderings were handled by the
electrophilic attack ( f k ) (Eq. (2)) and for radical attack ( f k0 ) (Eq. (3)), CYLview 1.0b software package [56].
respectively.
3. Results and discussion
f k+ = k (N + 1) k (N ) (1)
3.1. Assessing the level of theory
fk = k (N ) k (N 1) (2)
In order to assess the level of theory two different approaches have
k (N + 1) k (N 1)
f k0 = been used to reproduce the experimental UV–Vis spectra of the thiox­
2 (3)
anthone acetic acid based photoinitiators (Fig. 5).
where N is the number of electrons of a neutral molecule, k is the First, we have used TD-DFT with different functionals (B3LYP, M06-
electron density at the kth atomic position of neutral (N), anionic 2X and ωB97X-D) to understand which functional can reproduce the
(N + 1) and cationic (N − 1) species [50–54]. experimental wavelengths of thioxanthone acetic acid derivatives suc­
cessfully. Note that the all the optimized conformations generated with
2.5. Rate constants B3LYP have been re-optimized with M06-2X and ωB97X-D.
Subsequently the UV–Vis spectra of the optimized geometries have been
Reaction rate constants have been calculated by applying Eyring’s generated within the same methodology. It has been found that the

4
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Table 1
Excitation energies (eV), wavelength (nm), oscillator strength (f), nature, % contribution and experimental wavelength (nm) of thioxanthone acetic acid derivatives
(B3LYP/6-31+G(d,p) and 6-311++G(3df,3pd) for S atom in DMF).
Molecule eV Wavelength f Nature % Contribution Experimental Wavelength

TX-AA 3.3224 373.18 0.0928 HOMO → LUMO 96.57 384 [15]

TX-NH-AA 2.8476 435.39 0.0788 HOMO → LUMO 98.19 392–430 [16]
TX-O-AA 3.1809 389.78 0.0970 HOMO → LUMO 97.10 380–400 [17]
TX-S-AA 3.3062 375.01 0.0757 HOMO → LUMO 96.36 380–400 [17]

Fig. 6. Calculated and experimental UV–Vis absorption spectra of (a) TX-AA, (b) TX-NH-AA, (c) TX-O-AA, (d) TX-S-AA (B3LYP/6-31+G(d,p) and 6-311+
+G(3df,3pd) for S atom in DMF).

B3LYP/6-31+G(d,p) methodology reproduces the experimental λmax functionality [8,16]. TX-AA, TX-O-AA and TX-S-AA have shown large
values better than the M06-2X/6-31+G(d,p) and ωB97X-D/6- absorption maxima in the range of 380–400 nm [15–18]. Frontier or­
31+G(d,p) methodologies (Table 1, Tables S1-S4). After successfully bitals obtained through the vertical excitations are given in Fig. S1.
reproducing the experimental spectra with B3LYP/6-31+G(d,p), extra
basis set (6-311++G(3df,3pd)) for S atom has been added and in this 3.2. Natural transition orbital (NTO) calculations
way better agreement with experiment has been achieved (Table 1,
Fig. 6 and Tables S1-S4). Similarly, Wang et al. [25], Li et. al [26], Sun In order to understand the physical nature of electronic transitions
et al. [27], and Beni et al. [31,32] have reported the accuracy of B3LYP NTOs have been calculated and Φs values are reported. Φs is defined as
in their photophysical and structural investigations of thioxanthone a descriptor for light-induced electronic charge density distribution and
derivatives. obtained by analyzing the detachment and attachment density matrices
In the second part, the UV–Vis spectra are improved by Wigner [47,48]. The detachment can be described as the deficiency in electron
calculations in which the vibrational and dynamic effects are included density that results from light absorption. On the other hand, the at­
in the calculations (Fig. 6). Note that the simulated absorption char­ tachment can be described as an increase in electron density at the
acteristics match better the experimental data when vibrational effects excited state. Φs, without any dimension, can be expressed as the
are included by the Wigner distribution method. The single-conforma­ overlap between the attachment and the detachment densities. Φs can
tion dynamic approach taken into account by the Wigner distribution is have values in the range of 0.00 to 1.00 depending on the charge-
capable of reproducing the optical properties of organic chromophores transfer character of the electronic transition. If the Φs value is equal to
of different structures almost quantitatively [57–60]. 0.00, it indicates that there is no overlap between the attachment and
According to the reported experimental UV–Vis measurements; TX- detachment matrices. On the other hand, 1.00 means the transition
NH-AA has shown a tailing effect in the UV–Vis spectrum in the range results in zero photoinduced electronic density fluctuation. Φs values
392–430 nm [8,16]. Computationally, a corresponding peak has been close to 0.00 indicate charge-transfer transition whereas values close to
observed in the same region of the tailing effect (Fig. 6-b). The shift to 1.00 display a local excitation [47–49]. The occupied natural transition
longer wavelengths of TX-NH-AA, known as a panchromatic photo­ orbitals (oNTOs) represent the electron depletion density and the vir­
initiator, has been attributed to the hydrogen bonding ability of glycine tual natural transition orbitals (vNTOs) represent the electron

5
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Table 2
Occupied NTOs (oNTOs) and virtual NTOs (vNTOs) of S0 → S1 transition of thioxanthone acetic acid derivatives (B3LYP/6-31+G(d,p) and 6-311++G(3df,3pd) for S
atom in DMF).
Molecule Φs Value oNTO vNTO

TX-AA Φs-Löwdin = 0.701

Φs-Mulliken = 0.723

TX-NH-AA Φs-Löwdin = 0.695

Φs-Mulliken = 0.726

TX-O-AA Φs-Löwdin = 0.712

Φs-Mulliken = 0.732

TX-S-AA Φs-Löwdin = 0.725

Φs-Mulliken = 0.733

accumulation. Differently from the Kohn-Sham orbitals, representative important indicator of the efficiency of a photopolymerization system,
NTOs are capable of representing even complex excited states with a since it is the first step in the photopolymerization after the formation
single oNTO and vNTO couple [47–49]. of radicals [70–72]. In this study, we have modeled the radical attack to
The dominant transition in the UV–Vis spectra of thioxanthone the monomer both for exocyclic and endocyclic radicals of thiox­
acetic acid derivatives is S0 → S1 (Table 1 and Table 2). This transition anthone acetic acid derivatives and phenyl acetic acid derivatives
is also in good agreement with the experimental findings (Table 1 and (Fig. 8). Methyl methacrylate (MMA) has been used as the monomer
Fig. 6). The frontier orbitals indicate that all the local transitions of through all the modeled reactions both for thioxanthone acetic acid
thioxanthone acetic acid derivatives have a π-π* character (Fig. S1). derivatives and phenyl acetic acid derivatives, as reported in earlier
Note that the -NH, -O and -S between the thioxanthone core and the experimental studies [15–17].
acetic acid functionality have an influence on the oNTOs but not on the The activation barriers for TX-AAexo, TX-NH-AAexo, TX-O-AAexo, TX-
vNTOs. Table 2 reports the Φs-Löwdin and Φs-Mulliken values of the S-AAexo range between 15.30 and 18.72 kcal.mol−1 at B3LYP level and
dominant S0 → S1 transition of thioxanthone acetic acid derivatives. 15.44–19.34 kcal.mol−1 at BMK level, whereas the barriers for TX-
The Φs-Löwdin and Φs-Mulliken values are close to 1.00 and in each AAendo, TX-NH-AAendo, TX-O-AAendo, TX-S-AAendo are in the range
case the excitations can be attributed to local excitations (Table 2). 27.13–27.57 kcal.mol−1 at B3LYP level and 23.63–25.77 kcal.mol−1 at
BMK level. The reaction free energies calculated at B3LYP and BMK
3.3. Kinetic and thermodynamic parameters of the reactions levels show exergonic reaction behavior for all the reactions of exo­
cyclic radicals whereas the endocyclic radical reactions have an en­
It has been reported that in some cases B3LYP fails to treat open- dergonic reaction behavior (Table 4). The activation barriers of en­
shell systems, i.e. radicals, successfully and underestimate the reaction docyclic radicals are higher than the exocyclic radicals, thus at room
energies. This defect has been attributed to possible spin contamina­ temperature, the reactions are in favor of the exocyclic radicals and the
tion, self-interaction that arises in open-shell systems and over­ radical attack to the monomer of endocyclic radicals are unlikely to
stabilization of the structures. [42,44,61]. On the other hand, many occur. As previously experimentally proposed, endocyclic radicals may
authors claim the radicalic species to be treated successfully with lead to side product formation [15–17].
B3LYP [62–68]. As demonstrated above, B3LYP describes quite well the Even though thioxanthone acetic acid derivatives differ from each
experimental results obtained so far by Arsu et al. [15–17]. However, to other by the presence of the heteroatoms (-NH, -O and -S atom between
be on the safe side the BMK functional has been tested for the radicalic the thioxanthone core and the acetic acid functional group) endocyclic
reactions. The BMK functional is developed particularly for thermo­ radicals of thioxanthone acetic acid derivatives show similar properties
chemical kinetics and considers a higher weight in exact exchange, among themselves (Table 4). TXendo radical without the acetic acid
42%, compared to the B3LYP, which has 20% exact exchange functional group has similar kinetic and thermodynamic properties
[38–42,69]. with the endocyclic radicals of thioxanthone acetic acid derivatives. It
Overall radical formation reactions of thioxanthone acetic acid de­ is clear from this aspect that the side group functionality has no effect
rivatives and phenyl acetic acid based co-initiators (Fig. 7), where the on the kinetic and thermodynamic parameters of the initiation reaction
energy is provided by a light source, are endergonic as expected of endocyclic radicals.
(Table 3). Note that the same qualitative trend is followed with both func­
The radical attack to the monomer reaction (initiation) rate is an tionals (BMK and B3LYP) for the energetics of the radicals of

6
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Fig. 7. Overall radical formation reactions of (a) thioxanthone acetic acid derivatives and (b) phenyl acetic acid derivatives (X: -NH, -O, -S).

Table 3 Regarding the geometrical parameters of the transition state geo­

Gibbs free energies ( Grxn
o
kcal.mol−1) of overall radical formation reactions of metries of the thioxanthone acetic acid derivatives, the forming bond (d
thioxanthone acetic acid derivatives and phenyl acetic acid derivatives (B3LYP/ (CeC)) is 2.27, 2.41, 2.44 and 2.34 Å in the transition states of TX-
6-31+G(d,p) and 6-311++G(3df,3pd) for S atom in DMF, values in par­ AAexo, TX-NH-AAexo, TX-O-AAexo and TX-S-AAexo, respectively (Table 4
entheses are calculated with BMK/6-31+G(d,p) and 6-311++G(3df,3pd) for S and Fig. 9). For the endocyclic radicals TX-AAendo, TX-NH-AAendo, TX-
atom in DMF). O-AAendo and TX-S-AAendo, this distance is shorter: 2.11, 2.10, 2.11 and
Molecule o
Grxn Molecule o
Grxn 2.12 Å, respectively. These findings confirm Hammond’s postulate
[75–78] which states that shorter distances indicate a late transition
TX-AA 32.52 (41.89) PHC 33.07 (41.13) state with an endergonic character, whereas longer distances indicate
TX-NH-AA 34.73 (42.31) PHN 34.15 (40.93)
an early transition state (Fig. 9). The exocyclic radicals of phenyl acetic
TX-O-AA 40.16 (46.74) PHO 40.09 (46.42)
TX-S-AA 34.57 (43.87) PHS 35.71 (44.24) acid derivatives (PHCexo, PHNexo, PHOexo and PHSexo), which have si­
milar activation barriers as the thioxanthone acetic acid derivatives,
resemble in terms of the length of forming CeC bond as well (Table 4
thioxanthone acetic acid and phenyl acetic acid derivatives with the and Tables S14-S17).
exception of TX-O-AAexo. If necessary, even though computationally In the initiation reactions of exocyclic radicals of thioxanthone
demanding, composite methods such as CBS-QB3 [73,74] can be used acetic acid based species, a planar shape of the thioxanthone core has
to obtain accurate energies of molecules. been observed. However, for the endocyclic radical reactions of

Fig. 8. Radical attack to MMA of (a) exocyclic radical, (b) endocyclic radical of thioxanthone acetic acid derivatives, (c) exocyclic radical of phenyl acetic acid
derivatives and (d) endocyclic radical of thioxanthone (X: -NH, -O, -S).

7
E. Metin, et al. European Polymer Journal 136 (2020) 109909

Table 4 descriptors that can be used as a measure of the reactivity of a parti­

Activation free energy (ΔG‡ in kcal.mol−1), rate constant (k in L.mol−1.s−1), C- cular site on a molecule. Fukui functions allow the predictive approach
C distance (in Å) in transition state, overall free energy ( Grxn
o
in kcal.mol−1) of to define the most electrophilic, nucleophilic and radicalic sites on the
radical attack to monomer reactions (B3LYP/6-31+G(d,p) and 6-311+ molecule by using DFT [50–53]. Parr and Yang showed that the higher
+G(3df,3pd) for S atom in DMF, values in parentheses are calculated with the value of a corresponding Fukui function, higher the reactivity of
BMK/6-31+G(d,p) and 6-311++G(3df,3pd) for S atom in DMF).
that site of a molecule towards the corresponding attack [54].
Molecule G‡ k d(C-C) o
Grxn The spin densities and radical Fukui functions of the initiating ra­
dicals and the formed radicals (intermediates) after the radical attack to
TX-AAexo 18.72 (19.03) 2.88 2.27 −2.28 (−8.28)
the monomer (Fig. S2) have been calculated. The spin densities and
TX-NH-AAexo 15.30 (15.44) 9.27 × 102 2.41 −4.14 (−8.86)
TX-O-AAexo 15.60 (17.63) 5.58 × 102 2.44 −11.41 (−15.71)
radical Fukui function values of exocyclic radicals are similar both for
TX-S-AAexo 17.92 (19.34) 11.1 2.34 −6.40 (−11.11) thioxanthone and phenyl acetic acid derivatives (Table 5). However,
TX-AAendo 27.50 (23.63) 1.06 × 10−6 2.11 16.41 (7.96) spin densities and radical Fukui function values of endocyclic radicals
TX-NH-AAendo 27.13 (24.72) 1.97 × 10−6 2.10 16.39 (6.93) are significantly lower than the ones for the exocyclic radicals. Also
TX-O-AAendo 27.50 (25.77) 1.06 × 10−6 2.11 15.80 (7.70)
note that the spin density of the intermediate radicals of the endocyclic
TX-S-AAendo 27.57 (24.61) 9.39 × 10−7 2.12 16.60 (7.38)
PHCexo 18.61 (19.38) 3.47 2.27 −2.53 (−7.89) radical attack, is larger than its corresponding endocyclic radical
PHNexo 14.86 (16.42) 1.95 × 103 2.43 −4.81 (−9.54) (Table 5).
PHOexo 14.92 (16.64) 1.76 × 103 2.44 −11.83 (−14.83)
PHSexo 17.86 (18.63) 12.3 2.35 −7.54 (−13.56)
Table 5
TXendo 27.20 (24.27) 1.75 × 10−6 2.11 16.39 (7.76)
Spin densities and radical Fukui function values of initiating radicals and spin
densities of the intermediates formed in the radical attack to monomer reac­
tions (B3LYP/6-31+G(d,p) and 6-311++G(3df,3pd) for S atom in DMF).
Fukui Function Spin Density

Molecule f k0 Radical Intermediate

TX-AAexo 0.354 0.702 0.732

TX-NH-AAexo 0.393 0.760 0.719
TX-O-AAexo 0.454 0.881 0.736
TX-S-AAexo 0.386 0.787 0.734
TX-AAendo 0.197 0.433 0.721
TX-NH-AAendo 0.188 0.422 0.715
TX-O-AAendo 0.190 0.423 0.720
TX-S-AAendo 0.196 0.433 0.723
PHCexo 0.408 0.720 0.730
PHNexo 0.478 0.763 0.714
PHOexo 0.556 0.878 0.732
PHSexo 0.436 0.787 0.730
TXendo 0.197 0.433 0.721

We have carried out regression analyses between the spin densities

and radical Fukui function values vs. G‡ to correlate the spin densities
and/or the radical Fukui functions with the activation barriers (Fig. 10
and Fig. 11).
Including only thioxanthone based radicals have yielded R2 values
of 0.96 for spin density vs. G‡ and 0.97 for radical Fukui function
values vs. G‡ (Fig. 10). However, including phenyl acetic acid based

Fig. 9. Transition state geometries of thioxanthone acetic acid exocyclic and

endocyclic radicals (B3LYP/6-31+G(d,p) and 6-311++G(3df,3pd) for S atom
in DMF).

thioxanthone acetic acid based species, a non-planar bent butterfly

shape of the thioxanthone core has been observed; as this change in
structure was already reported earlier by Beni et al. [31]. The other
optimized ground state structures of the initiation reactions of exocyclic
and endocyclic radicals of thioxanthone acetic acid derivatives and
exocyclic radicals of phenyl acetic acid derivatives calculated with Fig. 10. Correlation between spin densities, radical Fukui functions ( f k0 ) and
B3LYP are given in the Tables S6-S18. G‡ of thioxanthone based initiating radicals (B3LYP/6-31+G(d,p) and 6-
Fukui function is one of the most well-known local reactivity 311++G(3df,3pd) for S atom in DMF).

8
E. Metin, et al. European Polymer Journal 136 (2020) 109909

since the available chemical knowledge related to the overall perfor­

mance of a designed photoinitiator could save energy and resources.

Declaration of Competing Interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influ­
ence the work reported in this paper.

Acknowledgements

Computing resources used in this work were partially provided by

the National Center for High Performance Computing of Turkey (UheM,
grant number 5005832018) and were partially performed at TUBITAK
ULAKBIM, High Performance and Grid Computing Center (TRUBA re­
sources).

Fig. 11. Correlation between spin densities, radical Fukui functions ( f k0 ) and Data Availability
G‡ of thioxanthone and phenyl acetic acid based initiating radicals together
(B3LYP/6-31+G(d,p) and 6-311++G(3df,3pd) for S atom in DMF).
The raw data required to reproduce these findings are available to
download from http://dx.doi.org/10.17632/stck8cf67v.2.
radicals have slightly lowered R2 values of 0.95 for spin density and
0.92 for radical Fukui function values vs. G‡ (Fig. 11). Appendix A. Supplementary material
Radical Fukui function values, spin densities and the corresponding
correlation diagrams calculated with BMK are given in Table S5 and Supplementary data to this article can be found online at https://
Figures S3-S4, respectively. doi.org/10.1016/j.eurpolymj.2020.109909.

References
4. Conclusion
[1] A. Palmeira, M.H. Vasconcelos, A. Paiva, M.X. Fernandes, M. Pinto, E. Sousa, Dual
The photophysical, kinetic and thermodynamic properties of the inhibitors of P-glycoprotein and tumor cell growth: (Re) discovering thioxanthones,
One-Component Type II thioxanthone acetic acid photoinitiators, which Biochem. Pharmacol. 83 (2012) 57–68, https://doi.org/10.1016/j.bcp.2011.10.
004.
have been synthesized over many years, have been determined. [2] S.A. Cınar, M.N. Guven, T.N. Eren, B. Cesur, M. Aleksanyan, B. Dedeoglu, N. Okte,
This study has demonstrated the successful usage of B3LYP/6- V. Aviyente, F. Morlet-Savary, J. Lalevée, D. Avci, Structure-reactivity relationships
31+G(d,p) with extra basis set (6-311++G(3df,3pd)) for S atom to of novel monomeric photoinitiators, J. Photochem. Photobiol. A Chem. 329 (2016)
77–87, https://doi.org/10.1016/j.jphotochem.2016.06.015.
represent the absorption spectra of thioxanthone based molecules. [3] E. Metin, G.S. Batibay, N. Arsu, In–situ formation of self-assembled Ag nanoclusters
Competition between exocyclic and endocyclic radicals of thiox­ on ct-DNA in the presence of 2-mercaptothioxanthone by using UV–vis light irra­
anthone acetic acid derivatives has been investigated in terms of kinetic diation, J. Photochem. Photobiol. A Chem. 356 (2018) 1–6, https://doi.org/10.
1016/j.jphotochem.2017.12.009.
and thermodynamic parameters. It has been found that the difference [4] J. Kreutzer, K. Kaya, Y. Yagci, Poly(propylene oxide)-thioxanthone as one-compo­
between the activation energies of the exocyclic radical attack and nent Type II polymeric photoinitiator for free radical polymerization with low
endocyclic radical attack to monomer is significant under the same migration behavior, Eur. Polym. J. 95 (2017) 71–81, https://doi.org/10.1016/j.
eurpolymj.2017.07.030.
reaction conditions. Thus, the attack of exocyclic radicals is favorable in
[5] J.P. Fouassier, J. Lalevée, Photoinitiators for Polymer Synthesis, Wiley VCH Verlag
each modeled reaction. This finding also supports the proposal of Arsu GmbH & Co. KGaA, Weinheim, 2012. doi:10.1002/9783527648245.
et al. [15–17] where they have proposed that exocyclic radicals are [6] N. Arsu, M. Aydin, Y. Yagci, S. Jockusch, N.J. Turro, One component thioxanthone
responsible for the initiation of the polymerization reactions. based Type II photoinitiators, Photochem. UV Curing New Trends. 661 (2006)
1–13.
The energetic trends obtained with BMK follow qualitatively the [7] Y. Yagci, S. Jockusch, N.J. Turro, Photoinitiated polymerization: Advances, chal­
ones calculated with B3LYP. However, if necessary, the structures al­ lenges, and opportunities, Macromolecules 43 (2010) 6245–6260, https://doi.org/
ready located with either one of the DFT functionals can be treated with 10.1021/ma1007545.
[8] S. Dadashi-Silab, C. Aydogan, Y. Yagci, Shining a light on an adaptable photo­
the composite CBS-QB3 method to obtain accurate energies [73,74]. initiator: advances in photopolymerizations initiated by thioxanthones, Polym.
We were also able to compare the One-Component Type II and Chem. 6 (2015) 6595–6615, https://doi.org/10.1039/c5py01004g.
conventional Type II systems. We have observed similar activation en­ [9] N. Karaca, N. Ocal, N. Arsu, S. Jockusch, Thioxanthone-benzothiophenes as pho­
toinitiator for free radical polymerization, J. Photochem. Photobiol. A Chem. 331
ergies, exergonic reaction behaviors, spin densities and radical Fukui (2016) 22–28, https://doi.org/10.1016/j.jphotochem.2016.01.017.
function values of the radicals in the initiation reactions. By taking into [10] Q. Wu, Y. Xiong, Q. Liang, H. Tang, Developing thioxanthone based visible pho­
consideration these similarities and the photophysical advantages of toinitiators for radical polymerization, RSC Adv. 4 (2014) 52324–52331, https://
doi.org/10.1039/c4ra07614a.
thioxanthone acetic acid derivatives such as the higher capability of
[11] J. Qiu, J. Wei, Thioxanthone photoinitiator containing polymerizable N-aromatic
light absorption in a wide range of light spectrum and having the same maleimide for photopolymerization, J. Polym. Res. 21 (2014), https://doi.org/10.
kinetic efficiency of radical attack to the monomer, we believe that 1007/s10965-014-0559-4.
[12] J. Kreutzer, Y. Yagci, One-component, double-chromophoric thioxanthone photo­
thioxanthone acetic acid derivatives can be good candidates for the
initiators for free radical polymerization, J. Polym. Sci. Part A Polym. Chem. 55
replacement of conventional Type II photoinitiators. (2017) 3475–3482, https://doi.org/10.1002/pola.28729.
Regression analyses have suggested a fast and predictive approach [13] T.N. Eren, N. Yasar, V. Aviyente, F. Morlet-Savary, B. Graff, J.P. Fouassier,
for the determination of the activation energies by only calculating the J. Lalevee, D. Avci, Photophysical and photochemical studies of novel thiox­
anthone-functionalized methacrylates through LED excitation, Macromol. Chem.
spin density and/or the radical Fukui function of the initiating radicals. Phys. 217 (2016) 1501–1512, https://doi.org/10.1002/macp.201600051.
In conclusion, we believe that we have successfully proposed a [14] J. Lalevée, N. Blanchard, M. Ali Tehfe, C. Fries, F. Morlet-Savary, D. Gigmes,
stepwise approach for the determination of the initiation efficiency of J.P. Fouassier, New thioxanthone and xanthone photoinitiators based on silyl ra­
dical chemistry, Polym. Chem. 2 (2011) 1077–1084, https://doi.org/10.1039/
thioxanthone acetic acid based One-Component Type II photoinitiators c0py00392a.
and this approach can be utilized before moving to wet-lab experiments

9
E. Metin, et al. European Polymer Journal 136 (2020) 109909

[15] D.S. Esen, G. Temel, D.K. Balta, X. Allonas, N. Arsu, One-component thioxanthone [4,3,2-de]benzothieno[2,3-b]quinolines of biological interest, Bioorg. Chem. 94
acetic acid derivative photoinitiator for free radical polymerization, Photochem. (2020) 103347, https://doi.org/10.1016/j.bioorg.2019.103347.
Photobiol. 90 (2014) 463–469, https://doi.org/10.1111/php.12218. [38] G. M. J. Frisch, W. Trucks, H.B.. Schlegel, G.E.. Scuseria, M.A.. Robb, J.R.
[16] H. Tar, D. Sevinc Esen, M. Aydin, C. Ley, N. Arsu, X. Allonas, Panchromatic type II Cheeseman, G. Scalmani, V.. Barone, B.. Mennucci, G.A. Petersson et al., Gaussian
photoinitiator for free radical polymerization based on thioxanthone derivative, 09 Revision E.01, Gaussian Inc. Wallingford CT. (2013).
Macromolecules 46 (2013) 3266–3272, https://doi.org/10.1021/ma302641d. [39] A.D. Becke, A new mixing of Hartree-Fock and local density-functional theories, J.
[17] M. Aydin, N. Arsu, Y. Yagci, One-component bimolecular photoinitiating systems, Chem. Phys. 98 (1993) 1372–1377, https://doi.org/10.1063/1.464304.
2a: Thioxanthone acetic acid derivatives as photoinitiators for free radical poly­ [40] A.D. Becke, Density-functional exchange approximation with correct asymptotic
merization, Macromol. Rapid Commun. 24 (2003) 718–723, https://doi.org/10. behaviour, Phys. Rev. A. 38 (1988) 3098–3100, https://doi.org/10.1103/
1002/marc.200300019. PhysRevA.38.3098.
[18] F. Karasu, M. Aydin, M.A. Kaya, D.K. Balta, N. Arsu, Determination of photo­ [41] C. Lee, W. Yang, R.G. Parr, Development of the Colle-Salvetti correlationenergy
initiated polymerization of multifunctional acrylates with acetic acid derivatives of formula into a functional of the electron density, Phys. Rev. B. 37 (1988) 785–789.
thioxanthone by RT-FTIR, Prog. Org. Coatings. 64 (2009) 1–4, https://doi.org/10. [42] A.D. Boese, J.M.L. Martin, Development of density functional for thermochemical
1016/j.porgcoat.2008.07.004. kinetics, J. Chem. Phys. 121 (2004) 3405–3416, https://doi.org/10.1063/1.
[19] L. Cokbaglan, N. Arsu, Y. Yagci, S. Jockusch, N.J. Turro, 2-Mercaptothioxanthone as 1774975.
a novel photoinitiator for free radical polymerization, Macromolecules 36 (2003) [43] Y. Zhao, D.G. Truhlar, The M06 suite of density functionals for main group ther­
2649–2653, https://doi.org/10.1021/ma0214613. mochemistry, thermochemical kinetics, noncovalent interactions, excited states,
[20] F. Karasu, N. Arsu, S. Jockusch, N.J. Turro, Thioxanthone hydroquinone-O, O′- and transition elements: Two new functionals and systematic testing of four M06-
diacetic acid: photoinitiator or photostabilizer? J. Org. Chem. 78 (2013) class functionals and 12 other function, Theor. Chem. Acc. 120 (2008) 215–241,
9161–9165, https://doi.org/10.1021/jo401386t. https://doi.org/10.1007/s00214-007-0310-x.
[21] T. Çeper, N. Arsu, Photochemically prepared gold/polymer nanocoatings: forma­ [44] J. Da Chai, M. Head-Gordon, Long-range corrected hybrid density functionals with
tion of gold mirror, Macromol. Chem. Phys. 218 (2017) 1–8, https://doi.org/10. damped atom-atom dispersion corrections, Phys. Chem. Chem. Phys. 10 (2008)
1002/macp.201700030. 6615–6620, https://doi.org/10.1039/b810189b.
[22] N. Ataci, N. Arsu, Studies of the binding mode of TXNHCH2COOH with calf thymus [45] M.D. Hanwell, D.E. Curtis, D.C. Lonie, T. Vandermeersch, E. Zurek, G.R. Hutchison,
DNA by spectroscopic methods, Spectrochim. Acta - Part A Mol. Biomol. Spectrosc. Avogadro: an advanced semantic chemical editor, visualization, and analysis plat­
169 (2016) 128–133, https://doi.org/10.1016/j.saa.2016.06.032. form, J. Cheminform. 4 (2012) 1–17, https://doi.org/10.1186/1758-2946-4-17.
[23] G.S. Batibay, O.T. Gunkara, N. Ocal, N. Arsu, In-situ photoinduced formation of [46] M. Barbatti, M. Ruckenbauer, F. Plasser, J. Pittner, G. Granucci, M. Persico,
self–assembled Ag NPs using POSS-TX as nano-photoinitiator in PEGMEA/PEGDA H. Lischka, Newton-X: A surface-hopping program for nonadiabatic molecular dy­
polymer matrix and creating self-wrinkled pattern, J. Photochem. Photobiol. A namics, Wiley Interdiscip. Rev. Comput. Mol. Sci. 4 (2014) 26–33, https://doi.org/
Chem. 359 (2018) 73–79, https://doi.org/10.1016/j.jphotochem.2018.03.024. 10.1002/wcms.1158.
[24] M. Adibelli, E. Ozcelik, G.S. Batibay, T.O. Arasoglu, N. Arsu, A facile and versatile [47] T. Etienne, X. Assfeld, A. Monari, New insight into the topology of excited states
route for preparation AgNp nanocomposite thin films via thiol-acrylate photo­ through detachment/attachment density matrices-based centroids of charge, J.
polymerization: Determination of antibacterial activity, Prog. Org. Coatings. 143 Chem. Theory Comput. 10 (2014) 3906–3914, https://doi.org/10.1021/ct500400s.
(2020) 105620, https://doi.org/10.1016/j.porgcoat.2020.105620. [48] T. Etienne, X. Assfeld, A. Monari, Toward a quantitative assessment of electronic
[25] Z. Wang, Y. Li, X. Cai, D. Chen, G. Xie, K.K. Liu, Y.C. Wu, C.C. Lo, A. Lien, Y. Cao, transitions“ charge-transfer character, J. Chem. Theory Comput. 10 (2014)
S.J. Su, Structure-Performance Investigation of Thioxanthone Derivatives for 3896–3905, https://doi.org/10.1021/ct5003994.
Developing Color Tunable Highly Efficient Thermally Activated Delayed [49] B. Dedeoglu, A. Monari, T. Etienne, V. Aviyente, A.S. Özen, Detection of nitroaro­
Fluorescence Emitters, ACS Appl. Mater. Interfaces. 8 (2016) 8627–8636, https:// matic explosives based on fluorescence quenching of silafluorene- and silole-con­
doi.org/10.1021/acsami.5b12559. taining polymers: A time-dependent density functional theory study, J. Phys. Chem.
[26] Y. Li, X.L. Li, D. Chen, X. Cai, G. Xie, Z. He, Y.C. Wu, A. Lien, Y. Cao, S.J. Su, Design C. 118 (2014) 23946–23953, https://doi.org/10.1021/jp505025t.
Strategy of Blue and Yellow Thermally Activated Delayed Fluorescence Emitters [50] Y. Ma, J. Liang, D. Zhao, Y.L. Chen, J. Shen, B. Xiong, Condensed Fukui function
and Their All-Fluorescence White OLEDs with External Quantum Efficiency beyond predicts innate C-H radical functionalization sites on multi-nitrogen containing
20%, Adv. Funct. Mater. 26 (2016) 6904–6912, https://doi.org/10.1002/adfm. fused arenes, RSC Adv. 4 (2014) 17262–17264, https://doi.org/10.1039/
201602507. c4ra01853b.
[27] H. Sun, Z. Hu, C. Zhong, X. Chen, Z. Sun, J.L. Brédas, Impact of Dielectric Constant [51] L.H.M. Huizar, C.H. Rios-Reyes, N.J. Olvera-Maturano, J. Robles, J.A. Rodriguez,
on the Singlet-Triplet Gap in Thermally Activated Delayed Fluorescence Materials, Chemical reactivity of quinclorac employing the HSAB local principle - Fukui
J. Phys. Chem. Lett. 8 (2017) 2393–2398, https://doi.org/10.1021/acs.jpclett. function, Open Chem. 13 (2015) 52–60, https://doi.org/10.1515/chem-2015-0008.
7b00688. [52] A. Bendjeddou, T. Abbaz, A. Gouasmia, D. Villemin, Molecular Structure, HOMO-
[28] Y.-C. Chen, Y.-T. Kuo, T.-H. Ho, Photo-polymerization properties of type-II photo­ LUMO, MEP and Fukui Function Analysis of Some TTF-donor Substituted Molecules
initiator systems based on 2-chlorohexaaryl biimidazole (o -Cl-HABI) and various N Using DFT (B3LYP) Calculations, Int. Res. J. Pure Appl. Chem. 12 (2016) 1–9,
-phenylglycine (NPG) derivatives, Photochem. Photobiol. Sci. 18 (2019) 190–197, https://doi.org/10.9734/IRJPAC/2016/27066.
https://doi.org/10.1039/C8PP00300A. [53] L.R. Domingo, M. Ríos-Gutiérrez, P. Pérez, Applications of the conceptual density
[29] A. Al Mousawi, P. Garra, M. Schmitt, J. Toufaily, T. Hamieh, B. Graff, J.P. Fouassier, functional theory indices to organic chemistry reactivity, Molecules 21 (2016),
F. Dumur, J. Lalevée, 3-Hydroxyflavone and N-Phenylglycine in High Performance https://doi.org/10.3390/molecules21060748.
Photoinitiating Systems for 3D Printing and Photocomposites Synthesis, [54] R.G. Parr, W. Yang, Density functional approach to the frontier-electron theory of
Macromolecules 51 (2018) 4633–4641, https://doi.org/10.1021/acs.macromol. chemical reactivity, J. Am. Chem. Soc. 106 (1984) 4049–4050, https://doi.org/10.
8b00979. 1021/ja00326a036.
[30] J. Zhang, J. Lalevée, X. Mou, F. Morlet-Savary, B. Graff, P. Xiao, N-Phenylglycine as [55] P. Atkins, J. De Paula, Atkins’s Physical Chemistry, 9th ed., Oxford University Press,
a Versatile Photoinitiator under Near-UV LED, Macromolecules 51 (2018) New York, 2010.
3767–3773, https://doi.org/10.1021/acs.macromol.8b00747. [56] C.Y. Legault, CYLview Visualization and analysis software for computational
[31] A.S. Beni, A.N. Chermahini, H. Sharghi, S.M. Monfared, MP2 DFT and ab initio chemistry 1.0b, Univ. Sherbrooke. (2009).
calculations on thioxanthone, Spectrochim. Acta - Part A Mol, Biomol. Spectrosc. 82 [57] O. Sengul, M. Marazzi, A. Monari, S. Catak, Photophysical properties of novel two-
(2011) 49–55, https://doi.org/10.1016/j.saa.2011.06.059. photon absorbing dyes: assessing their possible use for singlet oxygen generation, J.
[32] A.S. Beni, S.M. Monfared, DFT and MP2 calculations on new series of hydro­ Phys. Chem. C. 122 (2018) 16315–16324, https://doi.org/10.1021/acs.jpcc.
xythioxanthones, J. Mol. Struct. 1039 (2013) 8–21, https://doi.org/10.1016/j. 8b04824.
molstruc.2013.01.060. [58] C. García-Iriepa, M. Marazzi, Level of theory and solvent effects on DASA absorp­
[33] E.G. Leggesse, W.R. Tong, S. Nachimuthu, T.Y. Chen, J.C. Jiang, Theoretical study tion properties prediction: Comparing TD-DFT, CASPT2 and NEVPT2, Materials
on photochemistry of Irgacure 907, J. Photochem. Photobiol. A Chem. 347 (2017) (Basel). 10 (2017) 1–13, https://doi.org/10.3390/ma10091025.
78–85, https://doi.org/10.1016/j.jphotochem.2017.07.018. [59] H. Gattuso, A. Monari, M. Marazzi, Photophysics of chlorin e6: from one- and two-
[34] D.E. Fast, A. Lauer, J.P. Menzel, A.M. Kelterer, G. Gescheidt, C. Barner-Kowollik, photon absorption to fluorescence and phosphorescence, RSC Adv. 7 (2017)
Wavelength-Dependent Photochemistry of Oxime Ester Photoinitiators, 10992–10999, https://doi.org/10.1039/c6ra28616j.
Macromolecules 50 (2017) 1815–1823, https://doi.org/10.1021/acs.macromol. [60] H. Gattuso, E. Dumont, M. Marazzi, A. Monari, Two-photon-absorption DNA sen­
7b00089. sitization: Via solvated electron production: Unraveling photochemical pathways by
[35] K. Kaya, J. Kreutzer, Y. Yagci, A Charge-Transfer Complex of molecular modeling and simulation, Phys. Chem. Chem. Phys. 18 (2016)
Thioxanthonephenacyl Sulfonium Salt as a Visible-Light Photoinitiator for Free 18598–18606, https://doi.org/10.1039/c6cp02592g.
Radical and Cationic Polymerizations, ChemPhotoChem. 3 (2019) 1187–1192, [61] D.H. Ess, T.C. Cook, Unrestricted prescriptions for open-shell singlet diradicals:
https://doi.org/10.1002/cptc.201800217. Using economical Ab initio and density functional theory to calculate singlet-triplet
[36] J.D. Williams, M. Nakano, R. Gérardy, J.A. Rincón, Ó. De Frutos, C. Mateos, gaps and bond dissociation curves, J. Phys. Chem. A. 116 (2012) 4922–4929,
J.C.M. Monbaliu, C.O. Kappe, Finding the Perfect Match: A Combined https://doi.org/10.1021/jp300633j.
Computational and Experimental Study toward Efficient and Scalable [62] B. Aubry, D. Subervie, M. Lansalot, E. Bourgeat-Lami, B. Graff, F. Morlet-Savary,
Photosensitized [2 + 2] Cycloadditions in Flow, Org. Process Res. Dev. 23 (2019) C. Dietlin, J.P. Fouassier, E. Lacôte, J. Lalevée, A Second-Generation Chameleon N-
78–87, https://doi.org/10.1021/acs.oprd.8b00375. Heterocyclic Carbene-Borane Coinitiator for the Visible-Light Oxygen-Resistant
[37] N. Mokhtari Brikci-Nigassa, L. Nauton, P. Moreau, O. Mongin, R.E. Duval, L. Picot, Photopolymerization of Both Organic and Water-Compatible Resins,
V. Thiéry, M. Souab, B. Baratte, S. Ruchaud, S. Bach, R. Le Guevel, G. Bentabed- Macromolecules 51 (2018) 9730–9739, https://doi.org/10.1021/acs.macromol.
Ababsa, W. Erb, T. Roisnel, V. Dorcet, F. Mongin, Functionalization of 9-thiox­ 8b01669.
anthone at the 1-position: From arylamino derivatives to [1]benzo(thio)pyrano [63] A. Mukhopadhyay, L. Jacob, S. Venkataramani, Dehydro-oxazole, thiazole and

10
E. Metin, et al. European Polymer Journal 136 (2020) 109909

imidazole radicals: insights into the electronic structure, stability and reactivity Phys. Chem. A. 119 (2015) 8261–8268, https://doi.org/10.1021/acs.jpca.5b04464.
aspects, Phys. Chem. Chem. Phys. 19 (2017) 394–407, https://doi.org/10.1039/ [71] X. Allonas, J. Lalevée, F. Morlet-Savary, J.P. Fouassier, Understanding the reactivity
c6cp05677f. of photoinitiating systems for photopolymerization, Polimery/Polymers. 51 (2006)
[64] H. Zhang, Y. Shang, M. Li, H. Zhao, X. Wang, B. Han, Theoretical study on the 491–498.
reaction mechanism in the UV radiation cross-linking process of polyethylene, RSC [72] J. Lalevée, X. Allonas, J.P. Fouassier, Prediction of initiation efficiency in radical
Adv. 6 (2016) 110831–110839, https://doi.org/10.1039/c6ra24433e. photopolymerization: The case of benzoyl and isopropylketyl radicals,
[65] H. Zhang, Y. Shang, M. Li, H. Zhao, X. Wang, B. Han, Theoretical study on the Macromolecules 38 (2005) 4521–4524, https://doi.org/10.1021/ma050203d.
radical reaction mechanism in the cross-linking process of polyethylene, RSC Adv. 5 [73] J.A. Montgomery, M.J. Frisch, J.W. Ochterski, G.A. Petersson, A complete basis set
(2015) 90343–90353, https://doi.org/10.1039/c5ra16339k. model chemistry. VI. Use of density functional geometries and frequencies, J. Chem.
[66] M. Yamaji, S. Tojo, M. Fujitsuka, A. Sugimoto, T. Majima, Mesolysis Mechanisms of Phys. 110 (1999) 2822–2827, https://doi.org/10.1063/1.477924.
Aromatic Thioether Radical Anions Studied by Pulse Radiolysis and DFT [74] J.A. Montgomery, M.J. Frisch, J.W. Ochterski, G.A. Petersson, A complete basis set
Calculations, J. Org. Chem. 80 (2015) 7890–7895, https://doi.org/10.1021/acs.joc. model chemistry. VII. Use of the minimum population localization method, J.
5b00660. Chem. Phys. 112 (2000) 6532–6542, https://doi.org/10.1063/1.481224.
[67] A. Zhao, Z. Li, H. Wang, Acetone/Water as a new photoinitiating system for pho­ [75] G.S. Hammond, A Correlation of Reaction Rates, J. Am. Chem. Soc. 77 (1955)
tografting: A theoretical study, Polymer (Guildf). 51 (2010) 2099–2105, https:// 334–338, https://doi.org/10.1021/ja01607a027.
doi.org/10.1016/j.polymer.2010.02.050. [76] V. Fındık, I. Degirmenci, Ş. Çatak, V. Aviyente, Theoretical investigation of thiol-
[68] A.G. Leach, R. Wang, G. Erich Wohlhieter, S.I. Khan, M.E. Jung, K.N. Houk, ene click reactions: A DFT perspective, Eur. Polym. J. 110 (2019) 211–220, https://
Theoretical elucidation of kinetic and thermodynamic control of radical addition doi.org/10.1016/j.eurpolymj.2018.11.030.
regioselectivity, J. Am. Chem. Soc. 125 (2003) 4271–4278, https://doi.org/10. [77] O.A. Ramos Rodríguez, N.E. Magaña Vergara, J.P. Mojica Sánchez, M.T. Sumaya
1021/ja029342q. Martínez, Z. Gómez Sandoval, A. Cruz, Á. Ramos Organillo, Synthesis, crystal
[69] M.E. Segovia, K. Irving, O.N. Ventura, Density functional and chemical model study structure, antioxidant activity and dft study of 2-aryl-2,3-dihydro-4H-[1,3]thiazino
of the competition between methyl and hydrogen scission of propane and β-scission [3,2-a]benzimidazol-4-One, J. Mol. Struct. 1199 (2020) 1–11. doi: 10.1016/j.mol­
of the propyl radical, Theor. Chem. Acc. 132 (2013) 1–18, https://doi.org/10. struc.2019.127036.
1007/s00214-012-1301-0. [78] K.S. Kharnaior, M. Devi, R.H. Duncan Lyngdoh, Generation of C7H7+ cations with
[70] H. Takahashi, Y. Marushima, K. Tsuji, K. Shibuya, A. Kawai, Addition Rate isomerization reactions, Comput. Theor. Chem. (2016) 150–164, https://doi.org/
Constants of Phosphorus- and Carbon-Centered Radicals to Double Bond of 10.1016/j.comptc.2016.07.018.
Monomers as Studied by a Pulsed Electron Paramagnetic Resonance Method, J.

11