Visible-light photosensitized oxidation of α-terpinene using novel silica-supported sensitizers: Photooxygenation vs. photodehydrogenation

Article abstract of DOI:10.1016/j.jcat.2013.04.001

Three silica-immobilized organic photocatalysts, based on rose bengal (RB), antraquinone-2-carboxylic acid (ANT-COOH) and a new cyanoanthracene derivative (DBTP-COOH), were prepared and characterized. Their efficiency for the photooxidation of α-terpinene was compared to that of their soluble counterparts. In solution, the three sensitizers showed high quantum yield of singlet oxygen production. Significant autooxidation to p-cymene occurred in the absence of catalyst while the mechanism of the sensitized reaction strongly depended on sensitizer structure. With DBTP and RB materials, ascaridole was rapidly produced by singlet oxygen addition. In contrast, ANT-based sensitizers favored photodehydrogenation to p-cymene through an electron-transfer step inducing a radical chain reaction, followed by further p-cymene oxidation upon prolonged irradiation. The highest efficiency and selectivity were obtained for photooxygenation with DBTP-based materials, and for photodehydrogenation with ANT-based materials, these properties make them attractive for future applications as immobilized photocatalyst in solar synthesis, waste treatment, and microflow reactors.

Full text of DOI:10.1016/j.jcat.2013.04.001

Journal of Catalysis 303 (2013) 164–174
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Visible-light photosensitized oxidation of
a-terpinene using
novel silica-supported sensitizers: Photooxygenation vs. photodehydrogenation
Filippo Ronzani a, Nathalie Costarramone a, Sylvie Blanc a, Abdelhamid Khalil Benabbou a,b
,
Mickaël Le Bechec a, Thierry Pigot a, Michael Oelgemöller c, , Sylvie Lacombe a,
⇑
⇑
a Université de Pau et des Pays de l’Adour, CNRS UMR5254 – IPREM, Equipe Chimie-Physique, Hélioparc – 2 av. Président Angot, 64053 Pau Cedex 09, France
b Laboratoire Physico-Chimie des Matériaux, Catalyse et Environnement, Département de Chimie, Faculté des Sciences, Université des Sciences et de la Technologie d’Oran – B.P.
1505 El M’naouar, 31000 Oran, Algeria
c James Cook University, School of Pharmacy and Molecular Sciences, Townsville, QLD 4811, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Three silica-immobilized organic photocatalysts, based on rose bengal (RB), antraquinone-2-carboxylic
acid (ANT-COOH) and a new cyanoanthracene derivative (DBTP-COOH), were prepared and character-
ized. Their efficiency for the photooxidation of a-terpinene was compared to that of their soluble coun-
Received 28 February 2013
Revised 28 March 2013
Accepted 1 April 2013
terparts. In solution, the three sensitizers showed high quantum yield of singlet oxygen production.
Significant autooxidation to p-cymene occurred in the absence of catalyst while the mechanism of the
sensitized reaction strongly depended on sensitizer structure. With DBTP and RB materials, ascaridole
was rapidly produced by singlet oxygen addition. In contrast, ANT-based sensitizers favored photodehy-
drogenation to p-cymene through an electron-transfer step inducing a radical chain reaction, followed by
further p-cymene oxidation upon prolonged irradiation. The highest efficiency and selectivity were
obtained for photooxygenation with DBTP-based materials, and for photodehydrogenation with ANT-
based materials, these properties make them attractive for future applications as immobilized photocat-
alyst in solar synthesis, waste treatment, and microflow reactors.
Keywords:
Supported catalyst
Singlet oxygen
Electron transfer
Green chemistry
Photooxidation
Photooxygenation
Photodehydrogenation
Sensitizer
Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction
have subsequently linked these to various solid supports [46–48].
In this work, we compare the reactivity of different photosensitiz-
Photooxygenations with singlet oxygen (1O2) allow for the easy
and straight-forward synthesis of oxygenated compounds [1–5].
They are induced by catalytic amounts of a photosensitizer (Sens)
and use oxygen or air as a mild oxidant. Subsequently, photooxy-
genations have emerged as model transformations in green and so-
lar chemistry [6–11]. Synthetic photooxygenations have
furthermore been realized on technical scales [12–14] or under
continuous flow conditions [15–21]. Likewise, 1O2 has been used
successfully for depollution [22–28], disinfection [29–32], and
photodynamic therapy (PDT) [33–36]. A major disadvantage of
homogeneous photooxygenations is the need to remove the
soluble photosensitizer from the product mixture. To avoid this
drawback, solid-immobilized sensitizers have been developed
and have found widespread applications [37–45]. We have
recently established benzo[b]triphenylene-9,14-dicarbonitrile
(DBTP) and its derivatives as highly efficient 1O2 sensitizers and
ers (Scheme 1) in solution phase and supported on silica for the
photosensitized oxidation of -terpinene (1) [49–52]. The photo-
oxidation efficiencies of the novel silica-grafted sensitizers were
additionally compared to that of commercially available rose ben-
gal on polystyrene (RB–PSref) as a reference material [53–55].
a
2. Materials and methods
2.1. Materials and equipment
Rose bengal sodium salt (RB), anthraquinone-2-carboxylic acid
(ANT-COOH), N-hydroxysuccinimide (NHS, 97%), 1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide (EDC, >97%), p-cymene (3,
99%), 3-aminopropyl-functionalized silica gel (NH2ASi, 40–
63
l
m, 1 mmol gꢀ1 NH2 loading), 3-chloropropyl-functionalized
silica gel (ClASi, 230–400 mesh, 2.5% or 1 mmol gꢀ1 Cl loading),
rose bengal bound to polystyrene (RB–PSref
, 200–400 mesh,
0.1 mmol gꢀ1 loading), 5,6,11,12-tetraphenylnaphtacene (rubrene,
⇑
Corresponding authors. Fax: +61 7 4781 6078 (M. Oelgemöller), +33 5 5940
7622 (S. Lacombe).
P98%), and 1H-phenalen-1-one (perinaphthenone, PN, 97%) were
purchased from Sigma–Aldrich and used as received.
(P89%, Kosher/Halal certified) was purchased from Aldrich and
a-Terpinene
E-mail addresses: michael.oelgemoeller@jcu.edu.au (M. Oelgemöller), sylvie.
lacombe@univ-pau.fr (S. Lacombe).
0021-9517/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2013.04.001

F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
165
Scheme 1. Structures of photosensitizers (Sens) used.
was used without purification. GC–MS analysis revealed p-cymene,
-phellandrene, and cineole as main impurities. 9,14-Dic-
yanobenzo[b]triphenylene-3-carboxylic acid (DBTP-COOH) was
prepared as previously reported [47]. Solvents (anhydrous or spec-
trophotometric grade) were purchased from Sigma–Aldrich and
were used as received.
sensitizer’s molar amount or by the photon flux absorbed by the
photosensitizer, giving km (M minꢀ1 mmolSensꢀ1) and kP (M Ein-
steinꢀ1), respectively. The estimated values of the absorbed photon
flux were calculated using Eq. (2) [12,56], where Pa;k is the ab-
sorbed photon flux for a given wavelength, P0;k is the emitted pho-
ton flux for a given wavelength and Ak the absorbance of the
photosensitizer at that given wavelength (kabs > 300 nm, Pyrex
reactor).
a
The diffuse reflectance UV–Vis spectra (DRUV) of the powder
samples were measured with a Perkin–Elmer 860 Spectrophotom-
eter equipped with a 15 cm diameter integrating sphere bearing
the holder in the bottom horizontal position. They were recorded
at room temperature in steps of 1 nm, in the range 300–800 nm
with a bandwidth of 2 nm. The instrument was calibrated with a
certified Spectralon white standard (Labsphere, North Sutton,
USA). Diffuse Reflectance Infrared Fourier Transform (DRIFT, KBr)
spectra were measured on a Nicolet MAGNA-FTIR 560 Spectrome-
ter. Thermogravimetric analyses (TGA) were carried out with a
Netzsch STA 409 instrument under air flow (5 cm3 minꢀ1) using a
X
X
k
Pa;k
¼
P0;kð1 ꢀ 10ꢀA
Þ
ð2Þ
k
k
P0;k (Einstein sꢀ1 nmꢀ1) was estimated from spectroradiometric
data (W mꢀ2 nmꢀ1) by taking into account the photon energy at a
given wavelength and by normalizing to the Avogadro’s number
(NA). The total irradiances (P0) of the lamps emitting at 575 and
419 nm were 1.6 and 3.1 mW cmꢀ2, respectively. The values of irra-
diance were measured with an International Light ILT900 spectrora-
diometer, placing the light sensor at a fixed distance of 3 cm from
the source. The absorbances of the supported sensitizers were esti-
mated from their molar absorption coefficients in ACN and their
loading. The quantum yields of singlet oxygen production (UD) of
the photosensitizers under investigation in ACN solution were
determined using rubrene as a selective probe toward 1O2 and peri-
naphthenone (PN) as a reference sensitizer. The experimental setup
and the analytical method used have been described in detail earlier
[57–61].
heating rate of 5 °C minꢀ1
.
Irradiations were performed in a Rayonet reactor (RPR-200,
Southern New England Ultraviolet Company) equipped with
16 ꢁ 8 W lamps with kmax = 419 nm (RPR-4190 Å) for DBTP–Si
and ANT–Si, and kmax = 575 nm (RPR-5750 Å) for RB–PSref and
RB–Si, respectively. A constant flow of oxygen (Linde Oxygen
Instrument 5.0) was delivered by a Brooks Mass Flow Controller
5850 TR with Read Out & Control Electronics 0154. NMR spectra
were recorded on
a
Bruker Advance 400 spectrometer (1H,
400.13 MHz and 13C, 100.6 MHz) in CDCl3 at 25 °C using a 5 mm in-
verse broadband probe. GC–MS analysis was performed on a Perkin
Elmer Clarus 680 Gas Chromatograph (carrier gas: He 5.6–24 psi,
injector: T = 150 °C, method: 50–200 °C at 10 °C minꢀ1, 5 min at
200 °C) equipped with a Perkin–Elmer Elite-5 MS column (60 m,
2.2. General procedure for the synthesis of silica-supported
photosensitizers
DBTP–Si and ANT–Si were prepared by reacting DBTP-COOH
and ANT-COOH with NH2ASi. The given amounts (Table 1) of dried
DBTP-COOH or ANT-COOH were dissolved in 100 mL of a mixture
(90/10% v/v) of anhydrous dichloromethane and N,N0-dimethyl-
formamide (DMF) under argon. Once the solid was dissolved,
NHS was added and the reaction mixture was stirred for 1 h. EDC
was added, and the reaction mixture was left under stirring at
room temperature until thin layer chromatography indicated com-
plete conversion of free sensitizer. NH2ASi was added and the reac-
tion mixture was sonicated for 1 h and then filtered. The yellow
(DBTP–Si) or beige (ANT–Si) solids (Fig. SI-1) were washed by
Soxhlet extraction with tetrahydrofuran (THF) until no color was
visible by UV–Vis spectrophotometry.
0.25 mm ID, 0.25 lm df) and coupled with a Perkin Elmer Clarus
600S Mass Spectrometer (EI = 70 eV, scan m/z: 40–300, source:
T = 150 °C). Samples from the experiments with solid-supported
sensitizers were filtered through a plug of cotton prior to analysis.
The product yields were quantified according to the integration of
the corresponding chromatographic peaks relative to 1. UV–Vis
spectra in solution were obtained on a Varian Cary 5E UV–Vis–
NIR spectrophotometer using a Hellma UV Visible cell (0.01 cm
optical length).
For the photooxidation reactions, zero-order kinetics with re-
spect to 1 were observed under our conditions. The corresponding
integrated rate expression was used:
RB–Si was prepared according to the reported procedure of
Tamagaki and coworkers [62] by reacting commercial ClASi with
RB in 30 mL DMF under reflux for 17 h. After filtration and washing
of the colored silica with water, methanol, acetone and chloroform,
several Soxhlet extractions with methanol and chloroform were
carried out overnight until no visible color appeared in the effluent.
The amounts of reactants used for each material are listed in
Table 1, together with the amount of grafted sensitizer (calculated
½
a
-terpineneꢂ ¼ ½
a
-terpineneꢂ0 ꢀ kt
ð1Þ
A plot of [ -terpinene] vs. time yielded a straight line of slope k.
a
The concentrations were deduced from the absorbance measure-
ments using the Beer–Lambert law (l = 0.01 cm) at the absorption
maximum (267 nm) of
1
in acetonitrile (ACN, e267nm =
8160 Mꢀ1 cmꢀ1). Rate constants (k, M minꢀ1) were adjusted to
the data sets by linear regression and normalized by the

166
F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
Table 1
Experimental details for the synthesis of PS–Si and characteristic IR bands.
Sens
RB
Sens/g (mmol)
0.80 (0.79)
X–Sia/g (mmol-X)
4.0 (2.9)
NHS/g (mmol)
–
EDC/
–
l
L (mmol)
Sens-Si (mmol gꢀ1
0.19 (0.09)b
)
Grafting (%)
26
IR (KBr) (cmꢀ1
)
1481 (C@C)
1643 (C@O)
1541 (C@C)
1653 (C@O)
2216 (C„N)
1540 (C@C)
1653 (C@O)
1678 (C@O)
DBTP
ANT
0.057 (0.15)
0.12 (0.50)
1.0 (1.4)
3.7 (3.7)
0.028 (0.24)
0.13 (1.2)
100 (0.56)
129 (0.72)
0.14 (0.30)b
10
13
0.13 (0.34)b
a
Functionalized silica: X = NH2– (DBTP–Si and ANT–Si) and X = Cl– (RB–Si).
Sens loading calculated from TGA.
b
by UV spectroscopic analysis of the washings and by TGA) and the%
grafting (ratio between the potential number of matrix functional-
ities and the number of grafted ones in mmol).
at m/z 148, 105, and 91 were selected (see Figs. SI-2 and SI-7). In
almost all of the chromatograms, the peak of 2 (retention time, rt
14.74 min) was accompanied by a much smaller peak (<5%) at rt
15.73 min with an identical mass spectrum (Fig. SI-3). This peak
was attributed to isoascaridole 5 (Scheme 3b), a bisepoxide that
is known to arise from either photochemical or thermal isomeriza-
tion of 2 [50,63–66]. By changing the experimental conditions of
the GC–MS analysis (i.e., reducing the injector temperature to
120 °C), slightly lower amounts of 5 were determined, implying
that 5 is formed predominantly photochemically and to a lesser ex-
tent by thermal conversion of 2 in the GC injector. This assumption
was confirmed by irradiation of pure ascaridole with visible light,
confirming its slow isomerization (less than 10% after 6 h of irradi-
ation) to 5. The presence of 5 was also confirmed by 1H NMR spec-
troscopic analysis of the crude mixture and the characteristic pair
of doublets of the methine protons adjacent to the epoxide func-
tions was found between 3.1 and 3.3 ppm (in CDCl3) [65]. Since 5
unambiguously arises from 2, their combined yields are given.
The crude reaction mixture was evaporated carefully (630 °C
water bath to avoid decomposition [67–69]), and the oily residue
was analyzed by 1H NMR spectroscopy and again by GC–MS anal-
2.3. General procedure for photooxidation reactions
a
-Terpinene (267 mg) and the sensitizer (Table 3) were dis-
solved in 50 mL of acetonitrile (ACN) and loaded into a Pyrex flask
(kabs > 300 nm) equipped on its bottom with a sintered glass frit
and at its top with a gas outlet and a screw cap for sampling during
irradiation (Fig. 1). The reaction mixture was irradiated in a Rayon-
et reactor while stirring and purging with a gentle stream of oxy-
gen (16 mL minꢀ1) through the sintered frit. The reactor was
cooled by an internal cooling fan (ca. 30 °C). For solid-supported
photosensitizers, the same procedure was followed, but at the
end of the irradiation, the crude reaction mixture was filtered
and the solid residue was washed with fresh ACN (10 mL). Irradia-
tions were performed for 20–150 min until GC–MS monitoring
showed complete conversion. Reuse experiments were carried
out following the same experimental conditions. In these cases,
the photocatalyst was recovered each time, washed with ACN
and dried before reuse. Photooxygenation of p-cymene was con-
ducted under the same experimental conditions by simply substi-
tuting 1 with 3 (same molar amount).
ysis. Noteworthy, p-cymene and unreacted
a-terpinene were re-
moved with the solvent during evaporation. NMR analysis can
thus not be used to determine the composition of the crude prod-
uct [70]. Ascaridole was isolated by column chromatography on
silica gel using a gradient 10 ? 15 ? 20 vol.% ethyl acetate in
cyclohexane as a mobile phase. See SI (Session SI-1) for a complete
characterization of the main photoproducts.
The conversion of
a-terpinene was followed by the drop in
intensity of its UV band at 267 nm and by GC–MS analysis. Each
product was identified by comparison with reference spectra. Un-
der the chosen GC conditions, compounds 2 (ascaridole) and 4
(cuminaldehyde) showed very similar retention times. Simulta-
neous Collection of Full Scan and Selected Ion (SIFI) analysis was
therefore performed to determine their relative percentages. For
2, ions at m/z 168, 125, and 97 were selected (Fig. SI-2). For 4, ions
3. Results and discussion
3.1. Synthesis and characterization of the prepared materials
RB–Si was prepared starting from commercially available 3-
chloropropyl-functionalized silica gel and rose bengal sodium salt
(RB), following a one-pot procedure reported in the literature
(Scheme 2a) [62]. For the synthesis of DBTP–Si and ANT–Si, a
(a)
(b)
Fig. 1. Rayonet photoreactor setup: (a) schematic side view with inserted Pyrex
flask (1: sintered glass frit, 2: oxygen inlet tube, 3: oxygen outlet tube, 4: septum
port for sampling, 5: 8 W fluorescent tube) and (b) top view of actual setup with
16 ꢁ 8 W lamps.
Scheme 2. Synthesis of silica-supported photosensitizers (Sens–Si).

F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
167
(a)
(b)
Scheme 3. (a) Photosensitized oxygenation of
a
-terpinene (1). (b) Thermally/photochemically induced isomerization of ascaridole (2) to isoascaridole (5).
common peptide-coupling method using NHS and EDC was fol-
lowed [71–73]. When strictly excluding moisture from the reaction
medium, the desired coupling was readily achieved within 1 h and
by-products as well as unreacted sensitizer were removed easily
by washing (Scheme 2b). The sensitizer loading was subsequently
determined by UV spectroscopic analysis of the washings and by
TGA. The two methods gave different results but were in the same
order of magnitude for the amount of grafted sensitizer
(mmol gꢀ1). TGA determines the total amount of organic content
including the unreacted linker groups and the grafted moieties.
The values obtained from UV analysis, however, were closer to
the expected ones (on the basis of the initial amount of sensitizer
to be grafted). Consequently, these values were used for the
calculation of the amount of sensitizer under the operating condi-
tions for photooxidation experiments.
contrast, the DRUV spectrum of RB–Si revealed significant changes.
According to the original paper by Tamagaki and coworkers [62],
the color and intensity absorption of RB–Si varies depending on
the silica/RB ratio and hence on the amount of RB grafted. Com-
pared to solid RB and commercial RB–PSref, the spectrum of the or-
ange RB–Si showed a significant blue-shift with a broad absorption
The materials synthesized were subsequently characterized by
Diffuse Reflectance Infrared Fourier Transform Spectroscopy
(DRIFTS; see SI: Fig. SI-4). RB–Si showed typical bands at approx.
1481 cmꢀ1 and 1643 cmꢀ1, respectively, for
m(C@C) and m(C@O)
(a)
vibrations. The sharp and strong signal at approximately
1260 cmꢀ1, expected for an ether group in the final material, was
covered by the strong and broad signal from SiAO and SiACH2
stretching vibrations found in the IR spectra of all starting (un-
bound) silica gels and all the samples. DBTP–Si showed typical
bands at 1541, 1653 and 2216 cmꢀ1
(C@O) and
1540 cmꢀ1
,
correlating to
(C„N) vibrations, respectively. In addition to the
(C@C) vibration, ANT–Si showed two separate signals
m(C@C),
m
m
m
for C@O stretching (1653 and 1678 cmꢀ1), representing the amide
linker and the quinone of the anthracene skeleton (Table 1).
The solid-supported sensitizers were furthermore analyzed by
UV Diffuse Reflectance spectroscopy (DRUV), and the spectra are
displayed in Fig. 2 together with the emission spectra of the
relevant fluorescent tubes used for irradiations. The absorption
spectra of the unbound photosensitizers in methanol are shown
for comparison in Fig. 3. The DRUV spectrum of solid RB powder
showed a broad and unresolved band at about 610 nm compared
to its UV–Vis spectrum in solution. The main peak at 555 nm ob-
served in solution shifted slightly toward lower wavelengths and
the low-intensity shoulder at 520 nm strongly increased with a
hypsochromic shift of 20 nm. The DRUV spectrum of RB–PSref
was similar to that of the pure powder, but slightly shifted toward
higher wavelengths and gave a more intense peak at 560 nm. In
(b)
Fig. 2. (a) DRUV spectra of solid RB, RB–PSref, and RB–Si vs. irradiance of 575 nm
fluorescent tubes (secondary axis). (b) DRUV spectra of solid DBTP-COOH, DBTP–Si,
solid ANT-COOH and ANT–Si vs. irradiance of 419 nm fluorescent tubes (secondary
axis).

168
F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
Fig. 3. Electronic spectra of RB, DBTP-COOH, and ANT-COOH (secondary axis) in methanol at room temperature.
maximum between 450 and 570 nm (Fig. 2a). The DRUV spectra of
DBTP-COOH (powder) and DBTP–Si showed a broad, unresolved
absorption band between 450 and 380 nm (Fig. 2b). In contrast,
DBTP-COOH in solution displayed two well-resolved bands at
391 and 415 nm [46]. The DRUV spectra of ANT-COOH and ANT–
Si were consistent with the absorption spectrum of ANT-COOH in
solution with a very weak n ? pꢃ absorption band extending to
450 nm (Fig. 2b) [74,75].
Table 2
Quantum yields of singlet oxygen production of RB, DBTP-COOH, ANT-COOH
determined by chemical probing with rubrene (Reference Sens: perinaphthenone,
UD = 1.0. Solvent: air-equilibrated ACN at room temperature).
Sens
UD
RB
0.71 0.03
1.03 0.05
0.76 0.04
DBTP-COOH
ANT-COOH
The spectrophotometric analysis of the materials synthesized
confirmed that the peptide-coupling method used to graft DBTP-
COOH and ANT-COOH did not influence the absorption properties
of the immobilized dye relative to its pure powdered form. The loss
in fine structure relative to the solution spectra may be attributed
to the high sensitizer loading on the solid support and thus the for-
mation of aggregates, as already observed for silica-supported por-
phyrins [36], phthalocyanines [76], and methylene blue [77]. The
broad bands observed in the DRUV spectra of the solid-supported
and the powder sensitizers could also be explained by fluorescence
artifacts caused by the high concentration of the samples. The cou-
pling agents (NHS and EDC) and their reaction products are water-
soluble and were therefore easily removed together with the un-
grafted sensitizer. Thus, the chosen coupling method represents a
very efficient and simple way to graft carboxylic acid-derived
photosensitizers on commercially available silica gels.
3.3. Photosensitized oxidation of a-terpinene (1)
The silica-supported sensitizers and their homogeneous coun-
terparts were subsequently used for the photooxygenation of
a-terpinene (1) as a model reaction (Scheme 3a; Table 3)
[49–52]. Irradiation wavelengths were selected according to the
absorption maxima of the photosensitizers. Products and by-prod-
ucts were monitored by GC–MS to investigate their distribution
and were identified by their MS spectra. The decrease in the con-
centration of 1 was additionally followed by UV–Vis spectroscopy
without further dilution using 0.01 cm optical length cells. The
main products 2, 3, and 4 were furthermore confirmed by 1H
NMR analysis in comparison with literature data (SI, Session SI-1)
[39,63]. In case of the supported photosensitizers, subsequent UV–
Vis spectroscopic analysis of the final reaction mixture excluded
any leaching of the photosensitizer into the solvent.
A control irradiation for 2.5 h in the absence of any sensitizer at
575 nm (entry 1) resulted in autooxidation to p-cymene (3) as the
main product (39%). Other unidentified by-products (17%) mainly
contained two compounds with molecular ions of 168 and
152 a.m.u., respectively. Autooxidation thus represents an impor-
tant general pathway in the photooxidation of 1 and cannot be ne-
glected [64,78]. The addition of RB as a homogeneous sensitizer
(entry 2) clearly accelerated the reaction rate with complete con-
version of 1 after 30 min. Singlet oxygen addition to form 2 became
the dominant pathway and ascaridole was obtained in 78% yield,
together with 14% of 3 and 8% of other products. Likewise, UV–
Compared to the commercially available RB on polystyrene
(RB–PSref) at similar dye loading, RB grafting on ClASi under our
experimental conditions strongly altered the absorption spectrum
of the sensitizer. In this case, the formation of aggregates can be ru-
led out since this would correspond to a shift toward higher
wavelengths.
3.2. Determination of UD
Before performing the photooxygenation experiments, the
quantum yields of singlet oxygen production in ACN were estab-
lished for the soluble photosensitizers using the well-known
singlet oxygen probe, rubrene, and PN as a reference sensitizer
(Table 2) [57,58]. Among the three tested Sens, DBTP-COOH
showed the highest efficiency for 1O2 production with quantum
yield around unity. In contrast, RB and ANT-COOH have UD of
the same order of magnitude but about 25% lower than that of
DBTP-COOH. Thus, all chosen sensitizers produce singlet oxygen
very efficiently and selective photooxygenation of 1 to yield 2
was therefore expected.
Vis monitoring of the reaction showed a drop for the a-terpinene
band after 30 min (Fig. 4). The absorption band of RB underwent
a slight bathochromic shift and decreased in intensity. These ef-
fects probably resulted from a change in polarity of the reaction
medium, caused by the formation of 2 and 3, and photobleaching
(SI: Fig. SI-5) [79,80]. Prolonged irradiation over 2 h did not cause
any significant change in the product composition (Fig. 4). Rose

F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
169
Table 3
Experimental results for photooxygenations of 1.
Entry
Sens
Sens/mg (mmol)
kmax (nm)
t (min)
Composition (%)a
1
2b
3
4
Others
1
2
3
4
5
6
7
8
9
10
11
12
13
Nonec
RB
–
575
150
30
110
50
150
150
20
20
90
30
40
44
–
–
–
9
72
–
–
–
–
–
39
14
13
13
27
22
24
34
18
13
64
69
70
–
–
–
–
–
–
–
–
–
–
8
6
10
17
8
10
6
14
6
11
11
7
30.8 (0.030)
30 (0.003)
150 (0.015)
200 (0.038)
–
5.6 (0.015)
11.2 (0.030)
20 (0.003)
100 (0.014)
7.6 (0.030)
200 (0.030)
500 (0.065)
78
77
81
50
–
65
55
75
82
1
RB–PSref
RB–PSref
RB–Si
Nonec
419
DBTP-COOH
DBTP-COOH
DBTP–Si
DBTP–Si
ANT-COOH
ANT–Si
ANT–Si
5
–
–
–
27
22
16
110
80
3
4
a
b
c
Determined by GC analysis ( 1%; initial composition: 1 = 93%, 3 = 4%, others = 3%).
Including 5.
No sensitizer.
Fig. 4. (left) GC monitoring of 1, 2, 3, and others during the photooxygenation of 1 in the presence of RB (entry 2). (right) UV monitoring: intervals of 10 min, until
disappearance of the signal of 1 at 267 nm.
bengal thus prevented both further autooxidation of 3 and decom-
position of 2.
Photosensitized oxygenation with DBTP- and ANT-derived sen-
sitizers were carried out at 419 nm. In the absence of sensitizer
(entry 6), autooxidation was slow and 3 was found in an amount
of 22% after prolonged irradiation for 2.5 h, together with small
amounts of other products (total of 6%). The autooxidation is thus
slower at 419 nm than at 575 nm (entry 1). In the presence of sol-
uble DBTP-COOH (entries 7 and 8), very fast conversions of 1 were
achieved after 20 min. A further increase in DBTP-COOH concen-
tration was not considered since the selectivity toward 2 dropped
(possibly due to its flocculation caused by its low solubility in
ACN), while the amount of 3 increased from 24% to 34% (entry
8). DBTP-COOH nevertheless functions as an efficient singlet
oxygen sensitizer that does not catalyze any follow-up reactions.
In contrast, silica-supported DBTP-COOH (entries 9 and 10)
showed by far the best performance and resulted in rapid photoox-
ygenation and high selectivity toward 2. Reactivity improved with
higher photocatalyst loading (entry 10) and complete conversion
of 1 to 2 (82%) was achieved after 30 min of irradiation, a similar
result to that obtained with RB–PSref (entry 4). In addition, small
amounts of 3 (13–18%) and of other by-products (5–7%) were
formed. DBTP–Si reacted almost 1.5-times faster than RB–PSref at
In the presence of commercial RB–PSref (entries 3 and 4), similar
activity and selectivity were observed, although the reaction rate
depended on the molar amount of the grafted Sens. Using a low
loading of RB–PSref (entry 3), the reaction took 110 min to reach
completion, while with a higher loading (entry 4), full consump-
tion of 1 was reached after 50 min of irradiation, as confirmed by
GC as well as UV–Vis monitoring. Photoactivity sharply decreased
when RB–Si was employed as the photosensitizer (entry 5). After
prolonged irradiation for 2.5 h, the conversion was still incomplete
(91%) and selectivity for 2 was low (50%) while p-cymene (3)
was obtained in 27%, more than twice the amount found with
RB–PSref. The amount of other by-products likewise increased.
This poor activity and selectivity with RB–Si may be accounted
for by:
I. the poor overlap between the adsorption spectrum of RB–Si
and the emission spectrum of the fluorescent tubes (Fig. 2a)
compared to the much better overlap with the absorption
spectrum of RB–PSref
,
II. a structural change due to grafting, which could also explain
the orange color of the material and its modified DRUV spec-
trum and/or
III. the enhanced efficiency of RB–PSref due to swelling of the
organic PS-backbone in ACN, which may improve diffusion
of both O2 and 1 into the solid support [53–55].
a
comparable photosensitizer loading, while maintaining a
comparable selectivity. This effect may be due to the higher total
irradiance of the 419 nm tubes, in combination with the higher
UD for DBTP-COOH.
Due to the superior performance of DBTP–Si, a series of re-
peated experiments was conducted under standard conditions.

170
F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
Table 4
Reuse experiment: experimental results for photooxygenations of 1 with DBTP–Si
(irradiation times: 30 min).
Runa
Composition (%)b
1
2c
3
Others
1
2
3
4
1
1
1
1
86
81
82
83
6
8
7
8
7
10
10
8
a
Initial loading: 50 mg of DBTP–Si. Approximately 5 mg lost for each recovery.
Determined by GC analysis ( 1%; initial composition: 1 = 93%, 3 = 4%,
b
others = 3%).
c
Including 5.
Fig. 5. GC monitoring: 1, 2, 3, and others during the photooxygenation of 1 in the
presence of DBTP-COOH (entry 7).
As shown in Table 4, the recovered material performed equally
well even after four consecutive runs and the desired product 2
was obtained as main product in 81–86%. Selectivity was likewise
high and p-cymene was formed with just 6–8%. DBTP–Si therefore
represents an efficient, selective, and reusable photosensitizing
material for the catalytic photooxygenation of 1 by means of 1O2
addition.
observed, 4 was generated as the main product and its amount
constantly increased to 82% during the course of the irradiation.
The benzylic alcohols 7 and 8 were likewise rapidly detected, of
which p-cymene-7-ol (7) was further oxidized to 4 and 6 upon ex-
tended irradiation. When considering the combined yields of 4, 6
and 7, it may be concluded that oxidation preferentially took place
on the methyl group.
In case of anthraquinone-derived sensitizers (entries 11–13),
the photoreaction proceeded rather unselectively and 2 was only
obtained in trace amounts (1–4%). The reaction thus followed a dif-
ferent mechanism. After complete consumption of 1 (after 40–
110 min), p-cymene 3 was identified as the main photoproduct
(64–70%), together with smaller amounts of cuminaldehyde 4
(6–10%) and other products (16–27%; Scheme 4). Of these, euca-
lyptol (9) and isocineole (10) could be identified based on their
MS spectra. An additional product with a molecular mass of
168 a.m.u. was detected (Fig. SI-7), possibly the isomeric open
chain aldehydes 11 or 12, which were also identified in the crude
products by NMR spectroscopy [81,82]. The formation of 4 was fur-
thermore noticeable upon UV–Vis monitoring with the characteris-
3.4. Discussion of the mechanisms
Two different mechanistic schemes are deduced from our re-
sults. Soluble DBTP and RB, as well as RB–PSref and DBTP–Si photo-
sensitizers efficiently produced singlet oxygen and led selectively
to ascaridole 2 (including less than 5% of isoascaridole 5) via
[4 + 2] cycloaddition. Compound 2 is known to rearrange slowly
to yield 5, either upon direct irradiation or triplet sensitization
[85]. A slight increase of 5 with irradiation time was observed un-
der our experimental conditions. For pure 2, a slow isomerization
to 5 was independently observed (less than 10% in 6 h) with RB–
PSre. Any thermal isomerization can thus be excluded (though effi-
cient thermal isomerization of 2 to 5 has also been reported)
[50,65,67–69]. Besides, no significant changes in the product com-
positions were noticed upon prolonged irradiation. Compound 3,
commonly observed as a side-product from thermal dehydrogena-
tion (autooxidation) with ground state oxygen (3O2) [84], was only
observed in significant yields in the absence of sensitizer or with
RB–Si. Other mechanistic pathways to 3 (electron transfer involv-
ing the Sens [81,82], decomposition of ascaridole [86] or thermal
breakdown of hydroperoxide intermediates [87,88]) appear less
likely under our conditions.
tic
p ?
pꢃ transition at approximately 250 nm (Fig. 6) [83]. At
equal molar amounts of ANT-COOH (entry 11), the reaction rate
for the disappearance of 1 appeared to be in the same order of
magnitude as those of RB (entry 2) and DBTP-COOH (entry 8).
In contrast to the other sensitizers (Figs. 4 and 5), the product
composition further changed upon exhaustive irradiation and 3
was readily consumed, while the amounts of 4 and other by-prod-
ucts increased (Fig. 6). Among them, 4-tert-butylbenzoic acid (6)
could be verified by MS and NMR analysis (SI, Session SI-1), while
GC–MS analysis, in comparison with library spectra, additionally
suggested the presence of the isomeric benzylic alcohols 7 and 8,
together with the isomeric aldehydes 11 or 12 with molecular
mass 168 (Scheme 4; Fig. SI-7) [84].
The accelerating effect of anthraquinone on the secondary pho-
tooxidation of 3 was striking and was thus further investigated.
When 3 was independently irradiated for 160 min in the presence
of ANT-COOH, it was readily consumed (Fig. 7) and a variety of
oxygenated products were formed (Scheme 4). As previously
The mechanistic scenario that can explain the product distribu-
tion for the photooxidation of 1 in the presence of ANT-Sens is
outlined in Scheme 5 and mirrors the mechanism described by
Chen and coworkers for perylene diimides or 9,10-dicyanoanthra-
cene as photosensitizers [81,89]. Upon excitation and intersystem
Scheme 4. Identified secondary products obtained during the photooxygenation of
1 in the presence of ANT-Sens: 4-isopropylbenzaldehyde (cuminaldehyde, 4),
4-isopropylbenzoic acid (6), (4-isopropylbenzyl)methanol (p-cymene-7-ol, 7), 2-(4-methylphenyl)-2-propanol (p-cymene-8-ol, 8), 1,3,3-trimethyl-2-oxabicyclo[2.2.2]octane
(eucalyptol, 9), 1-isopropyl-4-methyl-7-oxabicyclo[2.2.1]heptane (isocineole, 10), and possible open chain aldehydes (3,7-dimethyl-2-octenal-6-on, 11) and (3-isopropyl-2-
heptenal-6-on, 12).

F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
171
Fig. 6. (left) GC monitoring of 1, 3, 4, and others during the photooxygenation of 1 in the presence of ANT-COOH (entry 11). (b) UV monitoring: measuring intervals of 10 min
(0 ? 150 min), until disappearance of the signal of 1 at 267 nm (gray) and the formation of 4, absorbing with a maximum at 250 nm.
Fig. 7. GC monitoring of 3 and 4 (principal axis), 6, 7, 8 and others (secondary axis)
during the photosensitized oxidation of 3 in the presence of ANT-COOH (homoge-
neous ACN solution, standard conditions).
crossing (ISC), the triplet excited state of the photosensitizer is
populated (Eq. (3)). Subsequent electron transfer (ET) between
the excited Sens and 1 generates a radical-ion pair (Eq. (4)), which
undergoes further proton transfer to the corresponding radical pair
(Eq. (5)). The resulting Sens-ANT-Hꢀ and 1 radicals react with
molecular oxygen to form the hydroperoxyl radical, HOOꢀ (Eqs.
(6) and (7)). This consequently undergoes hydrogen abstraction
from
1 to form hydrogen peroxide, thus regenerating the
a-terpinene-derived chain-carrier radical (Eq. (8)). Alternatively,
the anthraquinone radical anion can react with molecular oxygen
directly to generate the superoxide radical anion, Oꢀ2ꢀ (Eq. (9)),
which itself reacts via proton transfer with the radical cation 1ꢀ+
to create the hydroperoxyl radical, HOOꢀ (Eq. (10)). Likewise, the
Scheme 5. Plausible mechanism for the photooxidation of a-terpinene (1) with
ANT-derived sensitizers.
peroxidation of isomeric c-terpinene has been shown to yield p-
cymene as the only organic product in a chain reaction in which
the chain carrier is HOOꢀ [90,91]. The preferred oxidation of the
methyl group over the isopropyl group deduced from our results
strongly suggested an electron transfer rather than a free-radical
pathway [64,78,92].
A similar mechanism has been recently proposed by Cui and
coworkers [93] for the photooxidation of isopropanol with ANT-
COOH as a sensitizer: direct H-transfer between the photosensi-
tizer and the substrate generating a pair of radicals, O2 addition
on the substrate with formation of an unstable hydroperoxide
derivative, which deactivates by losing H2O2 to yield the oxidized
product.
already been proposed by Karapire and coworkers [94]. Under our
experimental conditions, the benzylic hydroperoxides [62,77]
could not be detected.
It follows from our results that singlet oxygen formation is
strongly disadvantaged with ANT and very favorable with RB
and DBTP sensitizers. From the Rehm–Weller equation, electron
transfer from
a-terpinene to the excited sensitizer (Eq. (4),
Scheme 5) should be exothermic either from the singlet or trip-
let excited ANT, and strongly endothermic with singlet or triplet
excited RB (Table 5). For DBTP, only the singlet state energy is
known [46] and electron transfer to this state should also be fa-
vored. However, it turns out that DBTP is at least as efficient as
The formation of 3, 12, and various by-products via the decom-
position of ascaridole and unstable benzylic hydroperoxides has

172
F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
Table 5
with the quantum yields of conversion of 1 (/conv), approximated
by calculating the ratio of the converted molar amount of 1 and
the photon flux absorbed by the sensitizer. The same trends are ob-
Calculated free energy for the electron-transfer reactions from
singlet or triplet excited states of the studied photosensitizers. Redox potentials are
a-terpinene to the
expressed vs. SCE for compounds in acetonitrile solutions.
D
ꢀ
G values obtained from
ꢀ DE00 Redox poten-
h
i
e02
aꢀ
the Rehm–Weller equation: DG ¼ 96:4 E0ðD=D ꢀ E0ðSens=Sens
tained considering kp or /conv
.
ꢀþ Þ
ꢀꢀ Þ
tial of 1: E0ð1=1 ¼ 1:60 V vs. SCE [101].
ꢀþ Þ
For the homogeneous solution experiments, the reactivity order
was ANT-COOH > DBTP-COOH > RB; the same trend was found for
the silica-supported sensitizers. Anthraquinone both in solution
and supported on silica was the most efficient photosensitizer,
with rate constants at least one order of magnitude higher than
for the other catalysts. DBTP, consistent to that observed in terms
of quantum yields of singlet oxygen production, was at least 1.5
times more efficient than RB. Notably, kp for DBTP–Si was only
20% lower than that for DBTP-COOH in homogeneous solution. In
contrast, the drops in activity for silica-supported RB and ANT-
COOH vs. their soluble forms were significant with about 40% for
RB–Si and 80% for ANT–Si, respectively. For RB-derived photosen-
sitizers, the rate constants strongly depended on the support and
RB–PSref was found four times more efficient than RB–Si. In addi-
tion, the rate constant did not increase consistently with higher
amounts of RB–PSref (entries 3 and 4): this behavior may be caused
by physical quenching of singlet oxygen by the solid sensitizer or
by unfavorable absorption of light due to scattering and diffraction
phenomena. For RB, silica was clearly not the most suitable sup-
port, while polystyrene was found advantageous for this reaction.
The kP and /conv values suggest that the reaction rates and effi-
ciencies depended on the mechanism. Compared to the DBTP- and
RB-derived sensitizers, ANT-based materials have a much higher
0
Sens
D
E00 (S1)
D
G (S1)
D
E00 (T1)
DG (T1)
ꢀꢀ
ESens=Sens
(kJ molꢀ1
)
(kJ molꢀ1
)
(kJ molꢀ1
)
(kJ molꢀ1
)
(V)
ANT
RB
ꢀ0.86
ꢀ1.20
[102]
284
213
ꢀ52.6
261
164
ꢀ26.7
55.0
100.1
DBTP ꢀ0.92
288
ꢀ50.8
–
–
RB for singlet oxygen production, without any evidence of elec-
tron transfer. Since the energy of the DBTP triplet state is not
easily determined, it is difficult to conclude on its properties.
With ANT, electron- or H-transfer are possible and further radi-
cal-chain mechanisms are favored on thermodynamic grounds:
D
Hf (reaction 7) = ꢀ31.9 kcal molꢀ1 and
DHf (reaction
8) = ꢀ18.8 kcal molꢀ1 [81]. This result confirms previous reactiv-
ity of ANT sensitizers with sulfides where high amounts of sul-
fonic and sulfuric acids were obtained and attributed to an
electron-transfer mechanism [75]. Moreover, although anthraqui-
none is a common photosensitizer, capable of producing singlet
oxygen by energy transfer upon irradiation (Table 2) [95,96],
anthraquinone-2-sulfonate was also shown to produce superox-
ide radical anion Oꢀ2ꢀ by electron transfer [95,96] in addition to
1O2 [97,98]. The question of anthraquinone sensitized superoxide
radical anion formation was raised [75,99,100], but should be
possible considering the reduction potential of ANT and O2
capability of photooxidizing a-terpinene (1): electron-transfer-in-
duced oxidative cleavage yielding 3 as a main product (Scheme 5)
must thus be much faster than photooxygenation to 2. For the
reactions involving 1O2 generation through energy transfer and
its cycloaddition to 1, the DBTP-based materials were at least twice
faster than their RB-derived counterparts, in line with what was
(E0AQ=AQꢀꢀ ¼ ꢀ0:86 V and EO0
¼ ꢀ0:57 V vs. SCE in ACN, respec-
2=Oꢀꢀ
2
tively) [99,100].
expected from the respective values of UD
.
As commonly observed with supported photocatalysts, the
reaction efficiency decreased: this may be due to higher physical
quenching of singlet oxygen by the suspended solid support.
3.5. Kinetic studies
Selected reaction rate constants (k) for the photosensitized oxi-
dation of 1 were determined by UV–Vis spectroscopy over the first
30 min of each experiment (Table 6), because in case of ANT-
derived sensitizers, the formation of strongly absorbing secondary
products, in particular 4, impacted on the values for longer
irradiation times. Under all experimental conditions, pseudo-
zero-order reactions were established. In addition, the reaction
rates were of the same order of magnitude.
4. Conclusions
The preparation, characterization, and use of novel silica-
immobilized organic photocatalysts based on rose bengal (RB),
antraquinone (ANT), and a newly developed cyanoanthracene
derivative (DBTP) were described. The efficiency of the solid-
The simple rate constant values (k) do not consider the different
absorption properties of the three sensitizers. In particular, their
molar absorption coefficients and the position of their absorption
maxima relative to the emission spectra of the lamps differ signif-
icantly (Figs. 2 and 3). Thus, approximate values of the rate con-
stants normalized by the photon flux absorbed by the sensitizer
(kP) were calculated and additionally compiled in Table 6, together
supported sensitizers for the photooxidation of a-terpinene 1
was compared to that of their soluble counterparts. In ACN, a high-
er singlet oxygen quantum yield (UD = 1.03) was measured for
DBTP than for RB and ANT (UD = 0.71 and 0.76, respectively)
relative to perinaphthenone. Although formation of ascaridole 2
by singlet oxygen addition to 1 has been extensively studied, the
fate of the by-products (followed by GC–MS and UV–Vis analysis)
Table 6
Kinetic constants for photosensitized conversion of 1 (k = kinetic constant directly calculated forms the evolution of UV absorption signals; kP = approximate values of kinetic
constants normalized by the estimation of the photon flux absorbed by the Sens; /conv = approximate values of quantum yields of conversion of 1).
Entrya
Sens
Sens (mmol)
k (10ꢀ2 M minꢀ1
)
kP (10ꢀ2 M Einsteinꢀ1
)
/
(10ꢀ6 mol1 Einsteinꢀ1
)
conv
2
3
4
5
RB
0.030
0.003
0.015
0.038
0.015
0.014
0.030
0.003
5.00 1.00
1.44 0.03
2.80 0.03
0.81 0.02
5.75 0.08
4.10 0.07
6.70 0.40
1.30 0.03
7.0 0.2
3.8 0.1
4.4 0.1
1.1 0.1
13.0 0.2
10.0 0.2
840 40
105
19
39
5
170
120
5
2
3
1
5
5
RB–PSref
RB–PSref
RB–Si
DBTP-COOH
DBTP–Si
ANT-COOH
ANT–Si
7
10
11
12
6300 100
880 15
180
4
a
Entry numbers refer to Table 2.

F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
173
[27] K. Ozoemena, N. Kuznetsova, T. Nyokong, J. Mol. Catal. A: Chem. 176 (2001)
29.
[28] R. Gerdes, D. Wöhrle, W. Spiller, G. Schneider, G. Schnurpfeil, G. Schulz-Ekloff,
J. Photochem. Photobiol., A 111 (1997) 65.
and of the reaction scheme appeared more complicated than ex-
pected and depended on the organic function of the immobilized
photocatalyst.
[29] A.K. Benabbou, C. Guillard, S. Pigeot-Rémy, C. Cantau, T. Pigot, P. Lejeune, Z.
Derriche, S. Lacombe, J. Photochem. Photobiol., A 219 (2011) 101.
[30] L. Ryskova, V. Buchta, R. Slezak, Cent. Eur. J. Biol. 5 (2010) 400.
[31] G. Jori, S.B. Brown, Photochem. Photobiol. Sci. 3 (2004) 403.
[32] M. Wainwright, J. Antimicrob. Chemother. 42 (1998) 13.
[33] M.I. Burguete, F. Galindo, R. Gavara, S.V. Luis, M. Moreno, P. Thomas, D.A.
Russell, Photochem. Photobiol. Sci. 8 (2009) 37.
[34] M.R. Hamblin, T. Hasan, Photochem. Photobiol. Sci. 3 (2004) 436.
[35] J. Moan, Q. Peng, Anticancer Res. 23 (2002) 3591.
[36] T.J. Dougherty, T.J. Dougherty, C.J. Gomer, G. Jori, D. Kessel, M. Korbelik, J.
Moan, Q. Peng, JNCI J. Natl. Cancer Inst. 90 (1998) 889.
[37] S. Lacombe, T. Pigot, in: Photochemistry, The Royal Society of Chemistry,
2010, pp. 307–329.
With DBTP- and RB-based sensitizers, ascaridole was the major
product. The DBTP-based material showed the highest efficiency
and selectivity for ascaridole which, coupled with its reusability,
makes it attractive for future applications.
On the other hand, ANT-based sensitizers favored selective
photodehydrogenation to p-cymene (3) followed by its oxidative
cleavage upon exhaustive irradiation time. In this case and in
agreement with previous literature data, an electron-transfer
mechanism inducing a radical chain reaction to p-cymene and its
oxidation products is proposed. ANT-derived materials are thus
inexpensive and easily prepared catalyst for photoinduced
electron-transfer reactions.
[38] M.I. Burguete, R. Gavara, F. Galindo, S.V. Luis, Catal. Commun. 11 (2010) 1081.
[39] S.M. Ribeiro, A.C. Serra, A.M.d’AR. Gonsalves, J. Mol. Catal. A: Chem. 326
(2010) 121.
We thus demonstrated the efficiency and selectivity of silica-
immobilized organic photocatalyst, prepared from commercial pre-
pared from commercial or readily available sensitizers. By suitable
choice of the organic part of the catalyst, naturally available com-
pounds can be selectively converted upon visible-light irradiation
and with oxygen as the sole reactant to either endoperoxides or aro-
matic derivatives, all of them known as high value products. All these
properties make these materials attractive for future green chemistry
applications as immobilized photocatalyst in solar synthesis [6–11],
waste treatment [22–32], and microflow reactors [15–21].
[40] D.T. Santos, J.Q. Albarelli, K. Joyce, M. Oelgemöller, J. Chem. Technol.
Biotechnol. 84 (2009) 1026.
[41] J. Wahlen, D.E. De Vos, P.A. Jacobs, P.L. Alsters, Adv. Synth. Catal. 346 (2004)
152.
[42] M. Benaglia, T. Danelli, F. Fabris, D. Sperandio, G. Pozzi, Org. Lett. 4 (2002)
4229.
[43] S. Ribeiro, A.C. Serra, A.M.d’A. Rocha Gonsalves, ChemCatChem 5 (2013)
134.
[44] M. Trytek, M. Majdan, A. Lipke, J. Fiedurek, J. Catal. 286 (2012) 193.
[45] S.M. Ribeiro, A.C. Serra, A.M.d’A. Rocha Gonsalves, J. Catal. 256 (2008) 331.
[46] S. Blanc, T. Pigot, C. Cugnet, R. Brown, S. Lacombe, Phys. Chem. Chem. Phys. 12
(2010) 11280.
[47] S. Lacombe, J.-P. Soumillion, A. El Kadib, T. Pigot, S. Blanc, R. Brown, E.
Oliveros, C. Cantau, P. Saint-Cricq, Langmuir 25 (2009) 11168.
[48] P. Saint-Cricq, T. Pigot, L. Nicole, C. Sanchez, S. Lacombe, Chem. Commun.
(2009) 5281.
Acknowledgments
[49] M.H. Spraul, S. Nitz, F. Drawert, Tetrahedron 47 (1991) 3037.
[50] G.O. Schenck, K.G. Kinkel, H.-J. Mertens, Justus Liebigs Ann. Chem. 584 (1953)
125.
This project was financially supported by the French National
Research Agency (ANR 10-BLAN-0803) and James Cook University
(CRIG 2011). We thank Renal BACKOV (CRPP, Bordeaux) for his
kind help with the TGA experiments.
[51] G.O. Schenck, Angew. Chem. 57 (1944) 101.
[52] G.O. Schenck, K. Ziegler, Naturwissenschaften 32 (1944) 157.
[53] J. Paczkowski, D.C. Neckers, Macromolecules 18 (1985) 1245.
[54] A.P. Schaap, A.L. Thayer, E.C. Blossey, D.C. Neckers, J. Am. Chem. Soc. 97 (1975)
3741.
[55] E.C. Blossey, D.C. Neckers, A.L. Thayer, A.P. Schaap, J. Am. Chem. Soc. 95 (1973)
5820.
[56] P.A. Hoijemberg, A. Chemtob, C. Croutxé-Barghorn, J. Poly, A.M. Braun,
Macromolecules 44 (2011) 8727.
[57] G. Günther, E. Lemp, A.L. Zanocco, J. Photochem. Photobiol., A 151 (2002) 1.
[58] F.M. Cabrerizo, M. Laura Dántola, G. Petroselli, A.L. Capparelli, A.H. Thomas,
A.M. Braun, C. Lorente, E. Oliveros, Photochem. Photobiol. 83 (2007) 526.
[59] C. Martí, O. Jürgens, O. Cuenca, M. Casals, S. Nonell, J. Photochem. Photobiol.,
A 97 (1996) 11.
[60] E. Oliveros, P. Suardi-Murasecco, T. Aminian-Saghafi, A.M. Braun, H.-J.
Hansen, Helv. Chim. Acta 74 (1991) 79.
[61] F. Wilkinson, W.P. Helman, A.B. Ross, J. Phys. Chem. Ref. Data 22 (1993) 113.
[62] S. Tamagaki, C.E. Liesner, D.C. Neckers, J. Org. Chem. 45 (1980) 1573.
[63] R.R.L. Martins, M.G.P.M.S. Neves, A.J.D. Silvestre, A.M.S. Silva, J.A.S. Cavaleiro,
J. Mol. Catal. A: Chem. 137 (1999) 41.
[64] H. Kropf, D. Wischer, Tetrahedron Lett. 10 (1969) 1751.
[65] J.V. Crivello, B. Yang, J. Polym. Sci., Part A: Polym. Chem. 33 (1995) 1881.
[66] R. Matush, G. Schmidt, Chem. Ztg. (1990) 382.
[67] L.F.R. Cafferata, R. Jeandupeux, R.S. Rimada, Acta Farm. Bonaerense (2005)
567.
[68] L.F.R. Cafferata, R. Jeandupeux, A.I. Canizo, Afinidad 2005 62(s.d.) 285.
[69] J. Boche, O. Runquist, J. Org. Chem. 33 (1968) 4285.
[70] A. Yavorskyy, O. Shvydkiv, C. Limburg, K. Nolan, Y.M.C. Delauré, M.
Oelgemöller, Green Chem. 14 (2012) 888.
[71] L. Zhao, K.M. Mullen, M.J. Chmielewski, A. Brown, N. Bampos, P.D. Beer, J.J.
Davis, New J. Chem. 33 (2009) 760.
[72] M. Storgaard, F.Z. Dörwald, B. Peschke, D. Tanner, J. Org. Chem. 74 (2009)
5032.
[73] J. Choithani, D. Sethi, P. Kumar, K.C. Gupta, Surf. Sci. 602 (2008) 2389.
[74] C. Cantau, S. Larribau, T. Pigot, M. Simon, M.T. Maurette, S. Lacombe, Catal.
Today 122 (2007) 27.
[75] V. Latour, T. Pigot, M. Simon, H. Cardy, S. Lacombe, Photochem. Photobiol. Sci.
4 (2005) 221.
[76] K. Ishii, Y. Kikukawa, M. Shiine, N. Kobayashi, T. Tsuru, Y. Sakai, A. Sakoda, Eur.
J. Inorg. Chem. 2008 (2008) 2975.
[77] E. Albiter, S. Alfaro, M.A. Valenzuela, Int. J. Photoen. 2012 (2012) 1.
[78] G.S. Serif, C.F. Hunt, A.N. Bourns, Can. J. Chem. 31 (1953) 1229.
[79] S.D.-M. Islam, O. Ito, J. Photochem. Photobiol., A 123 (1999) 53.
[80] M.A. Rauf, J.P. Graham, S.B. Bukallah, M.A.S. Al-Saedi, Spectrochim. Acta Part A
Mol. Biomol. Spectrosc. 72 (2009) 133.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2013.04.001.
References
[1] E.L. Clennan, A. Pace, Tetrahedron 61 (2005) 6665.
[2] M.R. Iesce, F. Cermola, F. Temussi, Curr. Org. Chem. 9 (2005) 109.
[3] M.C. DeRosa, R.J. Crutchley, Coord. Chem. Rev. 233 (2002) 351.
[4] E.L. Clennan, Tetrahedron 56 (2000) 9151.
[5] M. Prein, W. Adam, Angew. Chem., Int. Ed. 35 (1996) 477.
[6] N. Hoffmann, ChemSusChem 5 (2012) 352.
[7] M. Oelgemöller, C. Jung, J. Mattay, Pure Appl. Chem. 79 (2007) 1939.
[8] M. Oelgemöller, C. Jung, J. Ortner, J. Mattay, E. Zimmermann, Green Chem. 7
(2005) 35.
[9] A. Albini, M. Fagnoni, M. Mella, Pure Appl. Chem. 72 (2000) 1321.
[10] B. Pohlmann, H.-D. Scharf, U. Jarolimek, P. Mauermann, Sol. Energy 61 (1997)
159.
[11] P. Esser, B. Pohlmann, H.-D. Scharf, Angew. Chem., Int. Ed. 33 (1994) 2009.
[12] A.M. Braun, M.T. Maurette, E. Oliveros, Photochemical Technology, Wiley-
VCH, Weinheim, 1991.
[13] K. Gollnick, Chim. Ind. (1982) 156.
[14] W. Rojahn, H.-U. Warnecke, Dragoco-Report (1980) 159.
[15] F. Lévesque, P.H. Seeberger, Angew. Chem., Int. Ed. 51 (2012) 1706.
[16] O. Shvydkiv, C. Limburg, K. Nolan, M. Oelgemöller, J. Flow Chem. 2 (2012) 52.
[17] F. Lévesque, P.H. Seeberger, Org. Lett. 13 (2011) 5008.
[18] C.P. Park, R.A. Maurya, J.H. Lee, D.-P. Kim, Lab Chip 11 (2011) 1941.
[19] T. Carofiglio, P. Donnola, M. Maggini, M. Rossetto, E. Rossi, Adv. Synth. Catal.
350 (2008) 2815.
[20] K. Jähnisch, U. Dingerdissen, Chem. Eng. Technol. 28 (2005) 426.
[21] R.C.R. Wootton, R. Fortt, A.J. de Mello, Org. Process Res. Dev. 6 (2002) 187.
[22] M.L. Marin, L. Santos-Juanes, A. Arques, A.M. Amat, M.A. Miranda, Chem. Rev.
112 (2012) 1710.
[23] D. Gryglik, J.S. Miller, S. Ledakowicz, J. Hazard. Mater. 146 (2007) 502.
[24] J.S. Miller, Water Res. 39 (2005) 412.
[25] E. Pepe, O. Abbas, C. Rebufa, M. Simon, S. Lacombe, M. Julliard, J. Photochem.
Photobiol., A 170 (2005) 143.
[81] L. Chen, L.A. Lucia, E.R. Gaillard, D.G. Whitten, H. Icil, S. Icli, J. Phys. Chem. A
102 (1998) 9095.
[26] X. Tao, W. Ma, T. Zhang, J. Zhao, Chem. Eur. J. 8 (2002) 1321.

174
F. Ronzani et al. / Journal of Catalysis 303 (2013) 164–174
[82] C. Karapire, H. Kolancilar, Ü. Oyman, S. Icli, J. Photochem. Photobiol., A 153
(2002) 173.
[93] L. Cui, S. Furuhashi, Y. Tachikawa, N. Tada, T. Miura, A. Itoh, Tetrahedron Lett.
54 (2013) 162.
[83] A. Singh, V.K. Singh, M.A. Quraishi, Int. J. Corr. 2010 (2010) 1.
[84] J. Rudbäck, M.A. Bergström, A. Börje, U. Nilsson, A.-T. Karlberg, Chem. Res.
Toxicol. 25 (2012) 713.
[85] K.K. Maheshwari, P. de Mayo, D. Wiegand, Can. J. Chem. 48 (1970) 3265.
[86] W.G. Jary, T. Ganglberger, P. Pöchlauer, H. Falk, Monatsh. Chem. 136 (2005)
537.
[94] C. Karapire, M. Kus, G. Turkmen, C.C. Trevithick-Sutton, C.S. Foote, S. Icli, Sol.
Energy 78 (2005) 5.
[95] I. Gutiérrez, S.G. Bertolotti, M.A. Biasutti, A.T. Soltermann, N.A. García, Can. J.
Chem. 75 (1997) 423.
[96] V. Maurino, D. Borghesi, D. Vione, C. Minero, Photochem. Photobiol. Sci. 7
(2008) 321.
[87] R. Matusch, G. Schmidt, Helv. Chim. Acta 72 (1989) 51.
[88] R. Matusch, G. Schmidt, Angew. Chem. 100 (1988) 729.
[97] A.E. Alegrıa, A. Ferrer, G. Santiago, E. Sepúlveda, W. Flores, J. Photochem.
´
Photobiol., A 127 (1999) 57.
_
[89] S. Alp, Sß. Erten, C. Karapire, B. Köz, A.O. Doroshenko, S. Içli, J. Photochem.
[98] N. Liu, G. Sun, Ind. Eng. Chem. Res. 50 (2011) 5326.
[99] K. Gollnick, S. Held, J. Photochem. Photobiol., A 70 (1993) 135.
[100] T. Oppenländer, Photochemical Purification of Water and Air, Wiley-VCH,
Weinheim, 2003.
[101] N.B. Salah, F. Matoussi, J. Electroanal. Chem. 583 (2005) 1.
[102] A. Zakrzewski, D.C. Neckers, Tetrahedron 43 (1987) 4507.
Photobiol., A 135 (2000) 103.
[90] M.C. Foti, K.U. Ingold, J. Agric. Food Chem. 51 (2003) 2758.
[91] S. Sortino, S. Petralia, M.C. Foti, New J. Chem. 27 (2003) 1563.
[92] I. Saito, K. Tamoto, T. Matsuura, Tetrahedron Lett. 20 (1979) 2889.