Tailoring of organic dyes with oxidor eductive compounds to obtain photocyclic radical generator systems exhibiting photocatalytic behavior

Article abstract of DOI:10.3762/bjoc.10.92

The combination of a dye which absorbs the photon, an electron acceptor and an electron donor leading to energy conversion through electron transfer, was the basis of the so called three-component systems. In this paper, an experimental work combining Rose bengal dye with a triazine derivative as electron acceptor and ethyl 4-(dimethylamino)benzoate as electron donor, will underline the benefit of the photocyclic behavior of three-component systems leading to the dye regeneration. A thermodynamic approach of the photocycle is presented, followed by a mechanistic and computational study of ideal photocycles, in order to outline the specific kinetics occuring in so called photocatalytic systems. The simple kinetic model used is enough to outline the benefit of the cyclic system and to give the basic requirements in term of chemical combination needed to be fulfilled in order to obtain a photocatalytic behavior.

Full text of DOI:10.3762/bjoc.10.92

Tailoring of organic dyes with oxidoreductive
compounds to obtain photocyclic radical generator
systems exhibiting photocatalytic behavior
Christian Ley*, Julien Christmann, Ahmad Ibrahim, Luciano H. Di Stefano
and Xavier Allonas
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2014, 10, 936–947.
Laboratory of Macromolecular Photochemistry and Engineering,
University of Haute Alsace, 3b rue Alfred Werner, 68093 Mulhouse,
France
doi:10.3762/bjoc.10.92
Received: 22 January 2014
Accepted: 02 April 2014
Published: 25 April 2014
Email:
Christian Ley* - christian.ley@uha.fr
This article is part of the Thematic Series "Organic synthesis using
photoredox catalysis".
*
Corresponding author
Keywords:
Guest Editor: A. G. Griesbeck
computation; electron transfer; kinetic; photocatalysis;
photochemistry; photocyclic initiating system; photopolymerization;
photoredox catalysis; radical generator
© 2014 Ley et al; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
The combination of a dye which absorbs the photon, an electron acceptor and an electron donor leading to energy conversion
through electron transfer, was the basis of the so called three-component systems. In this paper, an experimental work combining
Rose bengal dye with a triazine derivative as electron acceptor and ethyl 4-(dimethylamino)benzoate as electron donor, will under-
line the benefit of the photocyclic behavior of three-component systems leading to the dye regeneration. A thermodynamic ap-
proach of the photocycle is presented, followed by a mechanistic and computational study of ideal photocycles, in order to outline
the specific kinetics occuring in so called photocatalytic systems. The simple kinetic model used is enough to outline the benefit of
the cyclic system and to give the basic requirements in term of chemical combination needed to be fulfilled in order to obtain a
photocatalytic behavior.
Introduction
Among the possible usage of light, the conversion of photons [6-8], and laser imaging [9]. In organic chemistry, the develop-
into chemical energy, as stored into radicals or ions, is of great ment of new methods for organic synthesis was achieved by use
interest. As a part of this research area, the development of of photolabile protective groups, which could be released under
photoradical generators (PRG) is still a lively topic that finds irradiation by light [10,11]. Since many decades in industry,
applications in triggering bioactivity [1], drug or fragrance PRG are used in photopolymerization, a field in which the PRG
release [2-4], microelectronics [5], water catalysis reduction prompt the initiation of the polymerization through a chain
936

Beilstein J. Org. Chem. 2014, 10, 936–947.
reaction [12]. Photopolymerization was first used over In this paper an experimental and mechanistic study of Type II
4
000 years ago in the mummification process [13]. During the PIS will be given and compared with a PCIS. Then, in order to
last decades the number of commercial applications is still improve the knowledge of PCIS, a thermodynamic and mecha-
continuously increasing. By example, photopolymer applica- nistic approach of PCIS exhibiting an ideal photocatalytic
tions are found in electronic materials [14], printing materials behavior will be presented. The proposed scheme will be used
[
15], optical and electro-optical materials [16,17], fabrication of as model to run some computation. This will permit to compare
devices and materials [18], adhesives and sealants [19], coat- and discuss the advantages, the specificity of this kind of photo-
ings [20] and surface modifications [21,22].
cyclic systems and outline the important features and condi-
tions which have to be fulfilled in order to obtain high perfor-
The great interest in PRG application to free radical polymeriza- mances for this kind of photocatalytic systems. This ideal ap-
tion (FRP) has led to the development of two major classes of proach will permit a better general understanding of the
photoinitiating systems (PIS): Type I and Type II. In Type I complex kinetics underlying PCIS chemical reactions,
PIS, the excited states reached after light absorption undergo a allowing a better and simplest selection of chemicals combina-
cleavage leading to the production of two initiating radicals tion.
[
5,23,24]. However, most Type I PIS are only active under
UV–blue irradiation [25,26]. To overcome this spectral limita- Results and Discussion
tion, Type II PIS were developed. They contain the photosensi- In order to outline this behavior, the photochemical consump-
tizer (PS) which absorbs the light and a coinitiator (Co) which tion of the dye (i.e., photolysis) was studied in a typical Type II
reacts with PS excited states through hydrogen abstraction or and a 3-cpt PCIS based on the Rose Bengal as dye, and a
electron transfer reaction (see Scheme 1). Numerous dyes were triazine derivative (TA) as an acceptor (coinitiator). In addition,
reported as PS [9,27-32]. Hydrogen donor coinitiators could be an amine (ethyl 4-(dimethylamino)benzoate, EDB) was chosen
amines [33-38], ethers [39,40], sulfides [41-43] or thiols [43- as redox (electron donor) additive for the PCIS (see Scheme 1).
4
5]. Electron transfer coinitiators could be borate salts [46,47],
iodonium [48,49] or triazine [32] derivatives. However, if Type II photoinitiating system
Type II PIS gain sensitivity in the visible part of the electro- In Scheme 2, the typical reaction mechanism of a dye (PS) with
magnetic spectrum, their efficiency is lower than Type I PIS. To an electron acceptor (A) is depicted. After absorption of actinic
gain more reactivity the so-called three component systems light (hν), the PS reaches its singlet excited state (1PS*) and
(
3-cpt) were developed by adding a redox additive to Type II after intersystem crossing (kISC), its triplet excited state (3PS*).
systems [32,50-52]. A higher yield of initiating radicals is gen- From both these excited states, an electron transfer can occur
erally claimed in such cases. Moreover, the dye is recovered from the PS to A (with quenching rate constants 1kq/A and
during the process and is newly available to absorb light, 3kq/A, respectively). The oxidized form of the dye (PS•+) is
running into a new cycle [9,32]. These photocyclic initiating formed together with the reduced acceptor (A•−) which lead to
systems (PCIS), should then present a somehow constant initiating radicals [33,48,49].
absorbance, leading to constant and efficient absorption of the
incident photons. Both features are responsible for the higher As a consequence, the PS is consumed (i.e., photolysed or
efficiency of PCIS [9,31,53-55] in photopolymerization reac- bleached) during the photoinduced electron transfer reaction.
tions. Therefore, as the dye is regenerated during the photo- Thus, according to Beer–Lambert's law, the absorbance of the
chemical reaction, a catalytic behavior appears, leading to the system decreases during the reaction: this bleaching could be
so-called photocatalytic system.
followed by UV–visible absorption spectroscopy.
Scheme 1: Chemical structures of RB, EDB and TA.
937

Beilstein J. Org. Chem. 2014, 10, 936–947.
To quantify the consumption of the PS, the photolysis quantum
yield (Φphotolysis) was determined. Φphotolysis is defined by the
ratio of the initial number of photosensitizer (PS) molecules
present (NPS) to the total number of absorbed photons (Nabs). It
could be expressed by the following equation:
(1)
where V is the volume of the irradiated solution. The absorbed
Scheme 2: Type II PIS mechanisms. PS: photosensitizer; 1,3PS*:
singlet and triplet PS excited states; PS•+: oxidized PS; A: electron
acceptor coinitiator; A•−: reduced electron acceptor.
photon concentration is given by:
(2)
The photolysis of an acetonitrile solution of RB/TA was done
within a 1 cm width cell with a monochromatic 532 nm laser
diode tuned to 9 mW/cm2 intensity. This will correspond to the where I0 is the incident light intensity on the cell (9 mW·cm−3,
mW/cm3 computation condition (vide infra). The initial i.e., 4·10−5 einstein·s−1·cm−3). The experimental measure of the
concentration was [RB]0 = 6.5 10−5 mol·L−1 (with ε = absorbance A(t) allows the calculation of Iabs(t) and a numer-
1900 M−1·cm−1 and a length of l = 1 cm, this corresponds to ical integration. Accordingly, the total concentration of
an initial absorbance of around 0.2 at 532 nm) and [TA]0 = absorbed photon is calculated as 3.43·10−4 mol·L−1 for RB/TA.
0−3 mol·L−1. The transmitted laser diode light was real-time It should be noted here that the missing first 1 second of the fast
9
3
1
recorded. It is thus possible to calculate the optical density A(t) decay of A(t) will represent, at maximum, 1.47·10−5 mol·L−1 of
of the sample, which is then converted into [RB](t). It can be absorbed photons (by assuming a constant A(t) = 0.2 during this
seen from Figure 1 that a fast decrease of [RB](t) occurs, so fast 1 s). This overestimation represents less than 5% of the total
that the very beginning is not resolved by the detector. This absorbed photons and could be neglected. Then, dividing the
indicates that the dye is consumed during irradiation, due to the initial RB concentration by these values, a photolysis quantum
electron transfer to TA. In about 100 s, the absorbance is almost yield of 0.19 for RB/TA is obtained.
zero indicating a complete consumption of the dye. As a conse-
quence, the absorption spectrum completely vanishes after Photocyclic initiating system
5
min of irradiation, confirming the disappearance of the dye A typical photocyclic initiating system (PCIS) consists of a
see insert Figure 1). It should be noted that in the same condi- light absorbing dye (PS), an electron acceptor (A) and an elec-
tions, no photolysis occurs for an acetonitrile solution of neat tron donor (D) (Scheme 3). In such systems, upon irradiation
(
RB.
photoinduced electron transfer reaction between the dye and
one of the components, for example A in Scheme 3, gives rise
to a radical anion (A•−) and the oxidized PS (PS•+). Then, PS•+
can react with the electron donor D (kred) to regenerate the PS
ground state (photocyclic reaction), leading to the formation of
one radical cation D•+. Generally, A•− and D•+ will give rise to
initiating radicals for free radical polymerization, to hydrogen
by water reduction or oxygen by water oxidation [6-8], etc. In
an ideal case, a photocatalytic behavior is ensured when there is
enough redox donor to make the dye surviving during a long
period of time. It will be shown below that this behavior has
great advantages; the most immediate is the fact that the
absorbance of the PCIS is kept constant.
In order to exemplify the behavior of the photocycle involving
the dye, the acceptor and the donor, the absorbance of a RB/TA/
EDB solution was monitored during irradiation as for the
Type II PIS. EDB was chosen as electron donor for the PCIS,
Figure 1: Evolution of RB concentration as a function of irradiation
time (λ = 532 nm, 9 mW·cm−3); insert: absorption spectra obtained
before and after irradiation.
938

Beilstein J. Org. Chem. 2014, 10, 936–947.
the absorbance slowly decreases due to an increase of the
photolysis rate with irradiation time. After 250 seconds, the
residual absorbance is zero: in the photocyclic system, the time
required for complete consumption of the dye is two times
longer than in Type II PIS. This confirms the dye regeneration
within a photocycle exhibiting a photocatalytic behavior.
The longer experimental surviving time of RB in the PCIS is in
qualitative agreement with the proposed schemes. To quantify
the cyclic behavior and to compare clearly Type II and PCIS,
the quantum yield of dye photolysis (Φphotolysis) was also deter-
mined for RB/TA/EDB. The total concentration of absorbed
photons for the PCIS is 0.0139 mol·L−1. Dividing the initial RB
concentration by this value leads to a photolysis quantum yield
of 4.7·10−3. This extremely low photolysis quantum yield
obtained for the PCIS means that more than 210 photons are
needed to bleach one dye molecule. The turnover number
becomes very high for the PCIS while for the Type II system,
only 5 photons are needed to bleach RB. This clearly demon-
strates the cyclic regeneration of the dye in the selected three-
component combination presented here. As a conclusion, these
Scheme 3: Photocatalytic behavior occurring in three component PIS.
PS: photosensitizer; 1,3PS*: singlet and triplet PS excited states;
A: electron acceptor; A•−: reduced form of acceptor; D: electron donor
and D•+: oxidized donor.
because its reactivity toward 3PS is low compared to that of TA experimental results clearly confirm the photocyclic behavior of
vide infra). This ensures that 3PS will react mainly with TA in the selected 3-cpt system RB/TA/EDB. However, to get more
(
an oxidative cycle as proposed in Scheme 3. The solution was insight and to understand the benefits of this process, a better
irradiated with the same laser diode and same output power. understanding of the thermodynamics and the kinetics is neces-
The same initial concentration of RB and TA was realized, and sary.
[
EDB]0 was fixed to 10−3 mol·L−1. The experimental evolution
of the [RB(t)] could be seen on Figure 2.
Thermodynamics of the PCIS
Obviously, every mixture of an acceptor, a donor and a dye,
would not give rise to a photocatalytic behavior. The compo-
nents should be selected with care in order to get a cyclic
behavior instead of competitive parallel reactions [54] where A
and D compete to react with the PS excited states. Thus, a ther-
modynamic approach of the PCIS should help the selection of
the candidates. Scheme 4 represents the different electron
transfer reaction occurring in the oxidative PCIS according to
the mechanism given in Scheme 3.
The first reaction is the electron transfer reaction between the
dye ground state and the coinitiator. Its reactivity is governed
by the corresponding Gibbs free energy change
where Eox and Ered are the half-wave oxi-
dation and reduction potentials for the donor and the acceptor,
respectively.
dark reaction.
must be as high as possible to prevent any
Figure 2: Evolution of [RB](t) in the photocyclic system RB/TA/EDB as
a function of irradiation time (λ = 532 nm, 9 mW·cm−3). Insert: absorp-
tion spectra obtained before and after irradiation.
After absorption of light the PS goes into singlet or triplet
excited states, in which it becomes both more oxidant and more
Compared to the Type II PIS, the behavior is quite different: the reducer. As a consequence, electron transfer reaction can occur
absorbance decreases very slowly at the beginning of the with the acceptor A (Scheme 3). The reaction must be as much
process, being almost constant for a couple of seconds. Then, exergonic as possible. The values of the Gibbs free energy
939

Beilstein J. Org. Chem. 2014, 10, 936–947.
Scheme 4: Thermodynamics of an oxidative three components PCIS, a) ground state reaction (
electron transfer (BET, ΔGBET), d) PS regeneration ( ).
), b) excited state reaction (
), c) back
change
for photoinduced electron transfer is given by the D should be carefully done to obtain an exergonic reduction of
Rehm–Weller equation [56]:
,
the oxidized PS.
where E* stands for the energy of the excited state. The
Coulombic term C is usually neglected in polar solvent.
This approach is also valid to a reductive catalytic cycle where
will determine the rate of electron transfer and the dye and the the PS first reacts with the donor D, and then its reduced form is
electron acceptor must be chosen such that is negative oxidized by the acceptor A. Table 1 summarizes the Gibbs free
enough to obtain a high electron transfer rate constant.
energy formula for estimation and the potential effect on the
photocyclic behavior. The "Wanted values" row gives the
The third electron transfer step is the unwanted back electron values which must be obtained in order to get a possible photo-
transfer BET within the contact ion pair (PS•+···A•−) which cyclic behavior of the selected components. The "Control" row
leads to initial reactants. This one is the more tricky to handle. summarizes the effect of the corresponding electron transfer on
ΔGBET is given by
. This reaction is gen- the PCIS final properties like shelf life, radical quantum yield
erally quite exergonic and the rate compete with the dissocia- and dye regeneration. One should also note that as
tion rate kdiss of the contact ion pair into free solvated species.
In practice BET reduce the overall radical generation quantum is wanted.
yields. Some new approaches allow to trigger this reaction
it is always negative even if a positive value
[
57,58].
Computational studies of PCIS
As seen before, there are four-electron transfer reactions to
The last important step is the dye regeneration. In order for the manage in order to get a working PCIS. Moreover the full
cycle and for the catalytic behavior to occur, the corresponding mechanistic description becomes more complicate by taking
Gibbs free energy change
assumes that the reduction potential of PS•+ is given by the oxi- be in competition (see Scheme 5). In order to get more insights
dation potential of PS, then . Thus, in order into the description and comprehension of PCIS, a complete
to achieve an efficient dye regeneration, the choice of the donor simulation of the photocyclic behavior of RB/TA/EDB was
should be negative. If one into account that both the oxidative and reductive pathways can
940

Beilstein J. Org. Chem. 2014, 10, 936–947.
Table 1: Gibbs free energy of the different electron transfer reactions occurring in a three component PCIS.
Gibbs free energy
Wanted values
Control
>0
<0
>0
<0
Shelf life
Rate of electron transfer: radical quantum yields
Control of back electron transfer: radical quantum yields
Regeneration of the PS: photolysis quantum yield
performed, comprising both a reductive and oxidative pathway
(
Scheme 5).
The full kinetic description of the reactions occurring in the
complete cycle is displayed in Figure 3, where Iabs is the
number of absorbed photons per second; kdes1, kdes3 are the rate
constants of singlet and triplet state deactivation to the ground
state, respectively; kISC the rate constant of the intersystem
crossing from 1PS to 3PS; 1kq/A, 3kq/A, (1kq/D, 3kq/D) the bimole-
cular electron transfer rate constant of the singlet and triplet
excited state of the PS by the acceptor A (donor D), respective-
ly; kred and kox are the rate constants of reduction and oxidation
of PS•+ and PS•−, respectively.
Scheme 5: General photocatalytic cycle occurring in three compo-
nents photocyclic systems. Two cycles are in competition: on the right
side the first reaction occurs with the donor D defining the reductive
pathway, while on the left side the primary reaction occurs with the
acceptor A leading to the oxidative pathway.
The time evolution of ground state PS 0PS, 1PS 3PS excited
state, PS•+, PS•−, D, A, D•+ and A•− are given by the following
equations:
Figure 3: Mechanistic description of photocyclic system involved in the RB/TA/EDB system. The rate constants are defined in the text.
941

Beilstein J. Org. Chem. 2014, 10, 936–947.
(3)
The absorption of the light is given by:
was taken to perform the computations. This is justified by the
fact that the radical recombination in the type II systems
observed by laser flash photolysis occurs in the ms timescale.
The last row of Table 2 contains the calculated Gibbs free
where I0 and IT are the incident and transmitted light intensity energy of the dye regeneration redox reaction. These slightly
on/through the sample respectively (einstein.L−1.s−1), ε is the endergonic values (0.42 and 0.12 eV) support the low kred and
molar extinction coefficient of the PS at the irradiation wave- kox rate constants used for computation.
length and l is the cell thickness.
The measured electron transfer rate constants are in line with
Simulation parameters
the ΔG, confirming the low reactivity of 3PS toward EDB.
In order to simulate the photocyclic behavior, the different rate Intersystem crossing rate constant kISC was obtained from the
constants involved in the proposed mechanism were measured triplet state quantum yield and singlet state lifetime according to
by time resolved spectroscopies (laser flash photolysis for [59]:
triplet excited states and time correlated single photon counting
for excited singlet states). The experimental quenching rate
constants and their corresponding ΔGET are given in Table 2.
As kred and kox were not measurable, a value of 2·103 M−1·s−1
Table 2: Thermodynamic and kinetic parameters of the PCIS simulation, kq: quenching rate constant of excited states.
Quencher
ΔGET (eV)/kq (M−1·s−1)
0RB
1RB
3RB
RB•+
RB•−
EDB
TA
2.07
1.77
−0.10/9.0 108
−0.40/6.0 109
0.27/4.50 104
0.42/–
−0.03/1.7 107
0.12/–
942

Beilstein J. Org. Chem. 2014, 10, 936–947.
which leads to:
and to
At this RB concentration, and in the absence of quencher, the
triplet state lifetime was measured around 80 µs, leading to
kdes3= 1.25·104 M−1s−1. The following parameters were used to
perform the computation: the continuous incident light inten-
sity I0 was fixed to 9 mW·cm−2 at 532 nm (i.e., 4·10−5
einstein·L−1·s−1). The initial PS (i.e., RB) concentration was
Figure 4: Evolution of RB, RB•+, TA, and TA•− concentration with time
for RB/TA system.
fixed to [RB]0 = 6.50·10−6 M, with ε = 31900 M−1·cm−1 and a equal to the initial ground state RB concentration, i.e.,
cell length of 1 cm this corresponds to an absorbance of 0.2 at 6.50·10−6 M. The final triazine concentration remains high
5
32 nm, in line with the experiments. The initial concentration because only 6.50·10−6 M of TA have reacted with the dye RB.
of the quenchers were fixed at [TA]0 = 10−2 M and [EDB]0 = The same conclusion can be drawn for the RB/EDB computa-
1
0−3 M for TA (acceptor A) and EDB (donor D).
tion but with slower rates due to lower reactivity (i.e., kq/D). As
a conclusion, in Type II PIS the limiting component for radical
generation is the concentration of the dye.
Photophysical cycle
One must keep in mind that in PCIS, two cycles can occur: the
photochemical one which involves the reaction of the dye Simulation of photocyclic initiating system
excited states with chemical reactants, and an internal photo- In the photocyclic system, the initial pseudo first-order reaction
physical. This later is an energy waste, comprising the absorp- rates are equal to 3kq/A[TA]0 = 1.7·105 s−1 for TA (acceptor A)
tion of light, the deactivation by internal conversion and fluo- and 3kq/D[EDB]0 = 4.5·101 s−1 for EDB (donor D). This means
rescence of 1PS to the ground state 0PS, the intersystem that the oxidative photocycle preferentially occurs, at least until
crossing to 3PS, and the deactivation of the 3PS to 0PS. In order 3kq/A[TA] > 3kq/D[EDB]. The plot of Log(3kq/A[TA]) and
to achieve high quantum yield for the conversion of the light Log(3kq/D[EDB]) on Figure 5 show that it is the case for the
energy, this photophysical cycle should be avoided as much as first 310 sec of the reaction: during this period one can assume
possible. This means that the excited states must live as long as that the oxidative pathway is the main reaction.
possible, the quenching rate constants should be as high as
possible (exergonic reaction with high 1,3kq/A and 1,3kq/D), and a
high concentration of D and A should be used (high pseudo first
order rate constants 1,3kq/A[A] and 1,3kq/D[D]). This will prevent
the probability of natural excited state deactivation to be high.
Simulation of Type II photoinitiating system
The Type II system studied here contains only RB as photosen-
sitizer and an electron acceptor TA (or donor EDB). Both
RB/TA and RB/EDB systems were calculated. Figure 4 shows
the changes in RB, RB•+, TA, and TA•− concentrations for the
former system.
Two important conclusions are to be outlined. First, the ground
state of the dye is quickly bleached: in about 4 seconds it has
completely disappeared. Second: the final TA•− concentration,
i.e., the maximum number of initiating radicals produced is
Figure 5: Evolution of RB, TA and EDB concentrations in the photo-
cyclic system. The logarithm of oxidative and reductive pseudo first
order reaction rates are given in plain and dashed lines, respectively.
943

Beilstein J. Org. Chem. 2014, 10, 936–947.
The evolution of [RB], [TA] and [EDB] displayed on Figure 5 photon absorption of the solution during the first part of the
is also very interesting. Two major differences with Type II PIS cycle;
should be outlined: first, the acceptor TA is totally consumed as
its concentration falls to zero. Second, the ground state RB 2. the final radical concentration is very high: 2.0065·10−3 M
concentration, presents only a slight decrease during 200 s, i.e., i.e. 150 times the concentration obtained in Type II PIS;
as long as enough acceptor TA is present in the solution. This
first part of the cycle should be explained by the reduction of 3. the limiting component is no more the dye, but the
the oxidized dye RB•+, by EDB leading to a recovery of the co-initiator of lowest concentration, in the present case the
dye. As soon as no more TA is available (around 250 s) [RB] acceptor TA.
lowers faster to zero. In this second part the photocatalytic
system is reduced to a conventional Type II system, where the All this phenomena explain the synergistic effect observed in
dye is consumed during its reaction with the excess of amine some free radical polymerization.
EDB. If the TA acceptor is completely consumed this is not the
case for the amine present in excess, only one TA equivalent is Radical and photolysis quantum yields
lost during the first part of the cycle, and [EDB] reduces to As for the Type II systems, in order to quantify the catalytic
0
6
.9·10−2 M. During the second part of the reaction, a more tiny behavior of the dye, the quantum yield of dye photolysis
.5·10−6 M is consumed (i.e. the initial RB concentration, cf. (Φphotolysis) in PCIS was determined. The radical quantum yield
type II PIS) leading to a final EDB concentration of is defined by the ratio of the number of produced radical to the
8
.9935·10−3 M.
total number of absorbed photons (Nabs). It could be expressed
by the following equation:
On Figure 6, the evolution of EDB•+ and TA•− concentrations
are displayed together with RB. In the first part of the cycle
(
before 300 s) both radical curves are similar. Then, as soon as
(4)
TA is consumed, no more TA•− is produced and a final concen-
tration of 1·10−3 M is reached, i.e. the initial TA concentration.
At this stage the EDB•+ also reaches 1·10−3 M. During this
second part, the reaction of excess EBD with RB leads to a tiny The calculated quantum yield for Type II RB/TA and photo-
more 6.5·10−6 M for EDB•+ and to the bleaching of the dye. cyclic RB/TA/EDB are given in Table 3.
Thus, the computed final total radical concentration is formally
equal to 2.0065·10−3 M.
Table 3: Calculated photolysis and radical quantum yields of the
Type II and the PCIS.
Type II RB/TA
PCIS RB/TA/EDB
Φphotolysis
Φrad
0.475
0.475
1.55·10−3
0.478
In Type II PIS both quantum yields are equal because the
amount of radicals formed is equal to the amount of dye photo-
lysed: as a consequence, only about two photons are needed to
bleach one dye molecule, corresponding to a very low turnover
number. Moreover we can see that the system has an average
efficiency: according to Φrad one photon on two is lost for
radical generation. For photocyclic system, the picture is
somehow different: the photolysis quantum yield falls to
Figure 6: Evolution of radical concentrations TA•− and EDB•+ together
with [RB] in photocatalytic system.
It is worth emphasizing the important points and advantages of 1.55·10−3, meaning that more than 600 photons are needed to
photocyclic systems:
bleach one dye molecule: the turnover number becomes very
high for the PCIS selected for the computation. However, if the
1
. the dye concentration is kept high for 250 s, this means that photocatalytic behavior is very good, the radical quantum yield
the absorbance of the solution is high, leading to a very high present an average value around 0.478, especially if one keep in
944

Beilstein J. Org. Chem. 2014, 10, 936–947.
mind that it could rise up to two, theoretically. But this is in line Steady sate UV–vis spectra were obtained on a Varian Cary
with the low rate constant of electron transfer of RB excited 4000 UV–vis double beam spectrophotometer in 1 cm path
singlet and triplet states: the wasting photophysical cycle is quartz UV grade cell.
comparatively too fast, preventing any efficient chemical
energy escape from the photophysical cycle
Laser flash photolysis experiments (LFP) were carried out
exciting at 532 nm with a nanosecond Nd-YAG laser (Power-
However one can see from the experimental results that the lite 9010, Continuum), operating at 10 Hz. The transient absorp-
consumption of the dye is longer than awaited for the Type II tion analysis system (LP900, Edinburgh Instruments) uses a
PIS, while it is shorter for the photocatalytic system. This is 450 W pulsed Xe arc lamp, a Czerny–Turner monochromator, a
confirmed by comparing the experimental photolysis quantum fast photomultiplier, and a transient digitizer (TDS 340,
yields to the computed values: the experimental photolysis Tektronix) [61]. The instrumental response was about 7 ns. The
quantum yield is higher and the turn over number of the real observation wavelength is indicated in each case. Experiments
photocatalytic system RB/TA/EDB is lower than expected, were performed in acetonitrile under Ar bubbling.
while the opposite stands for RB/TA. This should be due to the
fact that the bleaching of the dye is not only a direct conse- A FluoroMax-4 (Horiba, Jobin-Yvon) spectrofluorometer
quence of the electron transfer reaction but could be due also to coupled with a Time-Correlated Single-Photon Counting
secondary reactions between the radicals and the dye related (TCSPC) accessory was used to measure the steady-state fluo-
intermediates. As a consequence, the experimental efficiency of rescence spectra and singlet excited state lifetimes. NanoLEDs
the photocycle is lower than the simulated one. This is clear were used as pulsed excitation source leading to a time resolu-
from the shape of the two experimental curves where the tion of around 200 ps. The measurements were performed in
consumption of RB in RB/TA is initially fast and then decreases acetonitrile solutions under argon bubbling at room tempera-
slower while the opposite stands for the RB/TA/EDB: the ture. The quenching rate constants kq of the excited states were
consumption increases due to an increase of radical in the solu- obtained according to the Stern–Volmer analysis where the reci-
tion.
procal lifetime is plotted as a function of quencher concentra-
tion: [57].
Conclusion
In this paper experimental and full mechanistic studies of Chemicals: Rose Bengal extra (RB) and ethyl 4-(dimethyl-
Type II PIS and 3-cpt photocatalytic systems were presented amino)benzoate (EDB) were obtained from Aldrich, 2-(4-
and compared. Some advantages that PCIS bear over classical methoxyphenyl)-4,6-bis(trichloromethyl)-1,3,5-triazine (TA)
Type II systems are: dye regeneration with high possible was gifts from PCAS (France). Their chemical structures are
turnover (600), high radical production (>10−3 M). But these given in Scheme 1.
studies also reveal the importance of component choice: great
care must be taken in order to build photocatalytic radical References
generator by combination of three compounds. Especially a
simple thermodynamic approach should help to select the candi-
dates as a starting point. If the present paper was focused on
PCIS application for free radical photopolymerization, it should
be noted that the same type of system bearing the same under-
lying kinetics and mechanisms are used in the field of photocat-
alytic water reduction.
1. Wang, X.; Werner, S.; Weiß, T.; Liefeith, K.; Hoffmann, V. RSC Adv.
012, 2, 156–160. doi:10.1039/c1ra00599e
Muraoka, T.; Koh, C.-Y.; Cui, H.; Stupp, S. I. Angew. Chem. 2009, 121,
060–6063. doi:10.1002/ange.200901524
Muraoka, T.; Koh, C.-Y.; Cui, H.; Stupp, S. I. Angew. Chem., Int. Ed.
009, 48, 5946–5949. doi:10.1002/anie.200901524
2
2
.
.
6
3
2
4. Herrmann, A. Photochem. Photobiol. Sci. 2012, 11, 446–459.
doi:10.1039/c1pp05231d
5
.
.
Dietliker, K. A Compilation of Photoinitiators Commercially Available for
UV Today; SITA Technology Limited: Edinburgh/London, U.K., 2002.
Gärtner, F.; Denurra, S.; Losse, S.; Neubauer, A.; Boddien, A.;
Gopinathan, A.; Spannenberg, A.; Junge, H.; Lochbrunner, S.;
Blug, M.; Hoch, S.; Busse, J.; Gladiali, S.; Beller, M. Chem.–Eur. J.
2012, 18, 3220–3225. doi:10.1002/chem.201103670
Experimental
6
Redox potentials were measured by cyclic voltammetry using a
potentiostat (Princeton Applied Research 263A) at a scan rate
of 1 V/s in acetonitrile, with platinum as both working and
auxiliary electrodes, and a saturated calomel reference elec-
trode (KCl in methanol). Measurements were performed in
acetonitrile using 0.1 M of tetrabutylammonium hexafluoro-
phosphate (Aldrich) as supporting electrolyte. The samples
were bubbled with argon for 20 minutes prior to the analysis.
Ferrocene was used as standard [60].
7
.
.
Fukuzumi, S.; Hong, D.; Yamada, Y. J. Phys. Chem. Lett. 2013, 4,
3458–3467. doi:10.1021/jz401560x
8
Luo, S.-P.; Mejía, E.; Friedrich, A.; Pazidis, A.; Junge, H.;
Surkus, A.-E.; Jackstell, R.; Denurra, S.; Gladialli, S.; Lochbrunner, S.;
Beller, M. Angew. Chem., Int. Ed. 2013, 52, 419–423.
doi:10.1002/anie.201205915
945

Beilstein J. Org. Chem. 2014, 10, 936–947.
9.
Ibrahim, A.; Ley, C.; Allonas, X.; Tarzi, O. I.; Chan Yong, A.; Carré, C.;
Chevallier, R. Photochem. Photobiol. Sci. 2012, 11, 1682–1690.
doi:10.1039/c2pp25099c
32.Tarzi, O. I.; Allonas, X.; Ley, C.; Fouassier, J. P.
J. Polym. Sci., Part A: Polym. Chem. 2010, 48, 2594–2603.
doi:10.1002/pola.24039
1
0.Bochet, C. G. J. Chem. Soc., Perkin Trans. 1 2002, 125–142.
doi:10.1039/B009522M
33.Kucybaa, Z.; Pietrzak, M.; Paczkowski, J.; Linden, L.-A.; Rabek, J. F.
Polymer 1996, 37, 4585–4591. doi:10.1016/0032-3861(96)00302-3
34.Hageman, H. J. Prog. Org. Coat. 1985, 13, 123–150.
doi:10.1016/0033-0655(85)80021-2
1
1.Klán, P.; Šolomek, T.; Bochet, C. G.; Blanc, A.; Givens, R.; Rubina, M.;
Popik, V.; Kostikov, A.; Wirz, J. Chem. Rev. 2013, 113, 119–191.
doi:10.1021/cr300177k
35.Cook, W. D. Polymer 1992, 33, 600–609.
1
2.Fouassier, J. P.; Allonas, X.; Lalevee, J.; Dietlin, C. In Photochemistry
and Photophysics of Polymer Materials; Allen, N. S., Ed.; Wiley:
Hoboken, 2010; pp 351–419. doi:10.1002/9780470594179.ch10
3.Decker, C. J. Coat. Technol. 1987, 59, 97–106.
4.Peiffer, R. W. Applications of photopolymer technology. In
Photopolymerization: Fundamentals and Applications; Scranton, A. B.;
Bowman, C. N.; Peiffer, R. W., Eds.; American Chemical Society:
Washington, DC, USA, 1997; pp 1–14.
doi:10.1016/0032-3861(92)90738-I
36.Mateo, J. L.; Bosch, P.; Lozano, A. E. Macromolecules 1994, 27,
7794–7799. doi:10.1021/ma00104a038
1
1
37.Hoyle, C. E.; Kim, K.-J. J. Appl. Polym. Sci. 1987, 33, 2985–2996.
doi:10.1002/app.1987.070330831
38.Hoyle, C. E.; Keel, M.; Kim, K.-J. Polymer 1988, 29, 18–23.
doi:10.1016/0032-3861(88)90194-2
39.Buback, M.; Gilbert, R. G.; Russell, G. T.; Hill, D. J. T.; Moad, G.;
O’Driscoll, K. F.; Shen, J.; Winnik, M. A.
doi:10.1021/bk-1997-0673.ch001
1
5.Holman, R. Eur. Coat. J. 1995, 9, 610–612.
J. Polym. Sci., Part A: Polym. Chem. 1992, 30, 851–863.
doi:10.1002/pola.1992.080300516
1
6.Sponsler, M. B. J. Phys. Chem. 1995, 99, 9430–9436.
doi:10.1021/j100023a020
40.Soh, S. K.; Sundberg, D. C. J. Polym. Sci., Polym. Chem. Ed. 1982, 20,
1299–1313. doi:10.1002/pol.1982.170200513
1
7.Löchel, B.; Maciossek, A.; Quenzer, H. J.; Wanger, B.
J. Electrochem. Soc. 1996, 143, 237–244. doi:10.1149/1.1836415
8.Wakasa, K.; Chowdhury, N. A.; Priyawan, R.; Yoshida, Y.; Ikeda, A.;
Hirose, T.; Yamaki, M. J. Mater. Sci. Lett. 1996, 15, 134–136.
doi:10.1007/BF00291447
41.Andrzejewska, E. Polymer 1996, 37, 1039–1045.
doi:10.1016/0032-3861(96)87288-0
1
42.Guttenplan, J. B.; Cohen, S. G. J. Org. Chem. 1973, 38, 2001–2007.
doi:10.1021/jo00951a007
1
9.Shi, W.; Rånby, B. J. Appl. Polym. Sci. 1996, 59, 1951–1956.
doi:10.1002/(SICI)1097-4628(19960321)59:12<1951::AID-APP18>3.0.
CO;2-U
43.Lalevée, J.; Morlet-Savary, F.; El Roz, M.; Allonas, X.; Fouassier, J. P.
Macromol. Chem. Phys. 2009, 210, 311–319.
doi:10.1002/macp.200800566
2
0.Decker, C. Photostabilization of macromolecular materials by UV-cured
protective coatings. In Polymer Durability; Clough, R. L.;
Billingham, N. C.; Grillen, K. T., Eds.; American Chemical Society:
Washington, DC, USA, 1996; Vol. 249, pp 319–334.
doi:10.1021/ba-1996-0249.ch021
44.Hoyle, C. E.; Bowman, C. N. Angew. Chem., Int. Ed. 2010, 49,
1540–1573. doi:10.1002/anie.200903924
45.Morgan, C. R.; Kyle, D. R. UV Generated Oxygen Scavengers and
Method for Determining their Effectiveness in Photopolymerizable
Systems; Technology Marketing Corporation: Norwalk, CT, USA, 1983.
46.Kabatc, J.; Celmer, A. Polymer 2009, 50, 57–67.
doi:10.1016/j.polymer.2008.10.036
2
2
2
1.Akiyama, H.; Momose, M.; Ichimura, K.; Yamamura, S.
Macromolecules 1995, 28, 288–293. doi:10.1021/ma00105a040
2.Chan, C.-M.; Ko, T.-M.; Hiraoka, H. Surf. Sci. Rep. 1996, 24, 1–54.
doi:10.1016/0167-5729(96)80003-3
47.Kabatc, J.; Kucybała, Z.; Pietrzak, M.; Ścigalski, F.; Paczkowski, J.
Polymer 1999, 40, 735–745. doi:10.1016/S0032-3861(98)00282-1
48.Padon, S. K.; Scranton, A. B. J. Polym. Sci., Part A: Polym. Chem.
2000, 38, 3336–3346.
3.Allonas, X.; Morlet-Savary, F.; Lallevée, J.; Fouassier, J.-P.
Photochem. Photobiol. 2006, 82, 88–94.
doi:10.1562/2005-05-20-RA-535
doi:10.1002/1099-0518(20000915)38:18<3336::AID-POLA110>3.0.CO
;2-3
2
4.Jockusch, S.; Turro, N. J. J. Am. Chem. Soc. 1998, 120, 11773–11777.
doi:10.1021/ja982463z
49.Padon, K. S.; Scranton, A. B. J. Polym. Sci., Part A: Polym. Chem.
2000, 38, 2057–2066.
2
5.Crivello, J. V.; Dietliker, K. Photoinitiators for Free Radical Cationic &
Anionic Photopolymerisation; John Wiley & Sons: New York, USA,
doi:10.1002/(SICI)1099-0518(20000601)38:11<2057::AID-POLA140>3
.3.CO;2-X
1999.
2
2
2
2
3
3
6.Green, W. A. Industrial Photoinitiators: A Technical Guide; CRC Press:
50.Costela, A.; Costela, A.; Garcia-Moreno, I.; Sastre, R.
Macromol. Chem. Phys. 2003, 204, 2233–2239.
doi:10.1002/macp.200300010
Boca Raton FL, 2010. doi:10.1201/9781439827468
7.Allonas, X.; Fouassier, J. P.; Kaji, M.; Miyasaka, M.; Hidaka, T.
Polymer 2001, 42, 7627–7634. doi:10.1016/S0032-3861(01)00275-0
8.Mallavia, R.; Fimia, A.; García, C.; Sastre, R. J. Mod. Opt. 2001, 48,
51.Fouassier, J. P.; Allonas, X.; Burget, D. Prog. Org. Coat. 2003, 47,
16–36. doi:10.1016/S0300-9440(03)00011-0
9
41–945. doi:10.1080/09500340108230965
9.Mauguière-Guyonnet, F.; Burget, D.; Fouassier, J. P. Prog. Org. Coat.
007, 59, 37–45. doi:10.1016/j.porgcoat.2007.01.007
0.Mallavia, R.; Amat-Guerri, F.; Fimia, A.; Sastre, R. Macromolecules
994, 27, 2643–2646. doi:10.1021/ma00087a041
52.Allonas, X.; Fouassier, J. P.; Kaji, M.; Murakami, Y.
Photochem. Photobiol. Sci. 2003, 2, 224–229. doi:10.1039/b209203d
53.Grotzinger, C.; Burget, D.; Jacques, P.; Fouassier, J. P.
Macromol. Chem. Phys. 2001, 202, 3513–3522.
doi:10.1002/1521-3935(20011201)202:18<3513::AID-MACP3513>3.0.
CO;2-Z
2
1
1.Ibrahim, A.; Ley, C.; Tarzi, O. I.; Fouassier, J. P.; Allonas, X.
J. Photopolym. Sci. Technol. 2010, 23, 101–108.
doi:10.2494/photopolymer.23.101
54.Kabatc, J.; Jurek, K. Polymer 2012, 53, 1973–1980.
doi:10.1016/j.polymer.2012.03.027
55.Kabatc, J. Polymer 2010, 51, 5028–5036.
doi:10.1016/j.polymer.2010.09.020
946

Beilstein J. Org. Chem. 2014, 10, 936–947.
5
5
5
6.Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259–271.
doi:10.1002/ijch.197000029
7.Kawamura, K.; Schmitt, J.; Barnet, M.; Salmi, H.; Ley, C.; Allonas, X.
Chem.–Eur. J. 2013, 19, 12853–12858. doi:10.1002/chem.201300622
8.Kawamura, K.; Ley, C.; Schmitt, J.; Barnet, M.; Allonas, X.
J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 4325–4330.
doi:10.1002/pola.26844
5
9.Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern molecular
photochemistry of organic molecules; University science book:
Sausalito, Califormia, 2010.
6
0.Jacques, P.; Burget, D.; Allonas, X. New J. Chem. 1996, 30, 933.
1.Allonas, X.; Fouassier, J.-P.; Angiolini, L.; Caretti, D. Helv. Chim. Acta
6
2
001, 84, 2577–2588.
doi:10.1002/1522-2675(20010919)84:9<2577::AID-HLCA2577>3.0.CO
2-Q
;
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.10.92
947