Photocatalysis with Na4W10O32 in water system: Formation and reactivity of OH radicals

Article abstract of DOI:10.1016/j.molcata.2013.01.037

Photoexcitation of Na₄W₁₀O₃₂ dissolved in water leads to the formation of hydroxyl radicals. These species originate both from H₂O oxidation and H₂O₂ reduction. EPR-spin trapping investigation and laser flash photolysis experiments contribute to clarify their formation mechanism and their involvement in the oxidation of propan-2-ol. Continuous irradiation of water solutions of Na₄W ₁₀O₃₂ leads to the overoxidation of propan-2-ol to CO ₂ with high yield. This result may be of interest for the development of photocatalytic systems aimed to pollutants degradation. Entrapment of Na₄W₁₀O₃₂ into a microporous silica matrix gives a rather robust photocatalyst where the decatungstate structure is preserved. This material is able to catalyze the photooxidation of propan-2-ol to acetone with appreciable chemoselectivity. In particular, it is seen that the solid photocatalyst yields acetone as main product inhibiting its over-oxidation to carbon dioxide.

Full text of DOI:10.1016/j.molcata.2013.01.037

Journal of Molecular Catalysis A: Chemical 372 (2013) 23–28
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
•
Photocatalysis with Na W O in water system: Formation and reactivity of OH
4
10 32
radicals
a,∗
b
a
Alessandra Molinari , Roberto Argazzi , Andrea Maldotti
a
Dipartimento di Chimica, Università di Ferrara, Via Luigi Borsari 46, 44123 Ferrara, Italy
Centro ISOF CNR, Via Luigi Borsari 46, 44123 Ferrara, Italy
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 September 2012
Received in revised form
Photoexcitation of Na W10O32 dissolved in water leads to the formation of hydroxyl radicals. These
species originate both from H2O oxidation and H2O2 reduction. EPR-spin trapping investigation and
4
laser flash photolysis experiments contribute to clarify their formation mechanism and their involve-
ment in the oxidation of propan-2-ol. Continuous irradiation of water solutions of Na4W10O32 leads to
the overoxidation of propan-2-ol to CO2 with high yield. This result may be of interest for the devel-
opment of photocatalytic systems aimed to pollutants degradation. Entrapment of Na4W10O32 into a
microporous silica matrix gives a rather robust photocatalyst where the decatungstate structure is pre-
served. This material is able to catalyze the photooxidation of propan-2-ol to acetone with appreciable
chemoselectivity. In particular, it is seen that the solid photocatalyst yields acetone as main product
inhibiting its over-oxidation to carbon dioxide.
2
1 December 2012
Accepted 25 January 2013
Available online xxx
Keywords:
Sodium decatungstate
Heterogeneous photocatalysis
•
OH radicals
EPR-spin trapping
Alcohol oxidation
© 2013 Elsevier B.V. All rights reserved.
1
. Introduction
Tungsten oxygen anion clusters, known collectively as polyox-
an oxyradical-like character with a longer lifetime of about 50 ns.
wO is able to oxidize many organic substrates (RH) by hydrogen-
atom abstraction or electron transfer mechanisms, depending on
the chemical nature of RH. In both cases, the reaction pathways lead
otungstates (POT), form a large and structurally diverse class of
inorganic compounds with significant applications in photocata-
lysis [1–7]. In particular, their ability to promote photooxidation
reactions of various substrates has been investigated by several
authors [2,5–7]. This research topic moves toward a “sustainable
chemistry” opening to new oxidative routes either for syntheses or
for wastewater treatments that require O2 as available and cheap
oxidant, mild temperature and pressure conditions and use of light
5
−
to the one electron reduced form of the decatungstate (W10O32
)
•
and to the substrate derived radical (R , step c) [3,5,7,8,10]. Oxi-
dation of W10O32 by O restores starting W10O32 , closing the
5−
4−
2
photocatalytic cycle (step d). O2 consumption is accompanied by
its reductive activation to peroxy species [8,11].
In aqueous solution, the formation of the highly reactive
•
hydroxyl radicals (OH ) through the direct reaction of water
4−
as renewable source of energy. The decatungstate anion W10O32
with wO (Eq. (1)) is likely, since the excited state potential of
the decatungstate is more positive than the one-electron oxida-
tion of water [1]. However, this possibility is still a matter of
debate [17–20]. On the one hand, several kinetics reports sug-
gest that direct photoinduced electron/hydrogen transfer between
organic substrates and photoexcited decatungstate may occur also
is of particular interest from this point of view, since its absorp-
tion in the near UV region (ꢀmax = 323 nm) partially overlaps the UV
solar emission, opening the possibility to carry out solar-assisted
applications [1,2,7].
The proposed mechanism for W10O324 -based photocatalysis
−
•
in organic solvent involves absorption of light by the decatungstate
ground state leading to an oxygen to metal [O –W ] charge trans-
fer excited state (W10O32 , Scheme 1 step a) [8,9] that decays
in few picoseconds in both aerated and deaerated solutions to a
very reactive non-emissive transient wO (step b). This species has
in water without the involvement of OH radicals [19,20]. On
2−
6+
the other hand, some experimental observations are in agree-
ment with the formation of these intermediates. In particular,
it has been observed that illumination of aqueous solutions of
4−*
K W10O32 induces the hydroxylation of aromatic hydrocarbons, a
4
•
result consistent with OH -based mechanism but not exclusive to
it [17]. It has been also reported that photoexcitation of the poly-
i
∗ Corresponding author. Tel.: +39 0532455147; fax: +39 0532240709.
E-mail address: alessandra.molinari@unife.it (A. Molinari).
oxotungstate [NH3Pr ]5[W6O20(OH)] in aqueous solution in the
presence of the spin trap DMPO led to the evolution of an electron
1
381-1169/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2013.01.037

2
4
A. Molinari et al. / Journal of Molecular Catalysis A: Chemical 372 (2013) 23–28
of (nBu N) W10O32 in silica [14]. The obtained material con-
tained 30% (w/w) of decatungstate. UV–vis spectra of washing
4
4
hν
a)
4
-
4- *
W O
W O
10 32
1
0
32
water aliquots showed that Na W10O32 was not released into the
4
(
solution. A sample not including decatungstate (SiO ) was also
2
(
b)
prepared, following the same procedure mentioned above.
2
.2. EPR spin-trapping experiments
wO
RH
EPR spin-trapping experiments were carried out with a
Bruker ER200 MRD spectrometer equipped with a TE 201 res-
onator (microwave frequency of 9.4 GHz). The samples were
H O/CH CN (85/15) solutions containing 5,5-dimethylpyrroline N-
oxide (DMPO, 3 × 10 M) as spin trap, Na4W10O32 (2 × 10 M)
and, when requested, propan-2-ol (0.1 M). In heterogeneous exper-
iments, Na W10O32/SiO was suspended in the solution containing
(
c)
2
3
.
R + H+
−2
−3
(
d)
5
-
W O
1
0
32
4
2
DMPO and propan-2-ol as described above. Then, samples were put
into a flat quartz cell and directly irradiated into the EPR cavity at
+
ROOH
O + H
2
ꢀ
> 300 nm with a Hg medium pressure mercury lamp. No EPR sig-
Scheme 1. Photocatalytic behavior of W10O324 upon aerobic conditions.
−
nals were obtained in the dark and during irradiation of the solution
in the absence of photocatalyst.
paramagnetic resonance spectrum similar to that obtained from
•
the reaction between this spin trap and OH [21].
2.3. Laser flash photolysis experiments
wO + H O → W10O325 + OH• + H+
−
(1)
2
Nanosecond transient measurements were performed with a
custom laser spectrometer comprised of a Continuum Surelite II
Nd:YAG laser (FWHM 6–8 ns) with frequency doubled, (532 nm,
330 mJ) or tripled, (355 nm, 160 mJ) option, an Applied Photo-
physics xenon light source including a mod. 720 150 W lamp
housing, a mod. 620 power controlled lamp supply and a mod.
The partial oxidation of alcohols by O is a reaction with a high
synthetic value [12] and it is one of the most widely investigated
2
4
−
W10O32 -photocatalyzed reactions. Considering that this reaction
has been extensively studied in organic solvent, namely in acetoni-
trile solutions, we believe that an effort should be done in order to
investigate this process in water, which is a benign and econom-
ically advantageous solvent able to minimize the environmental
impact of organic synthesis. Herein, we report evidences of the for-
◦
03–102 arc lamp pulser. Laser excitation was provided at 90
with respect to the white light probe beam. Light transmitted
by the sample was focused onto the entrance slit of a 300 mm
focal length Acton SpectraPro 2300i triple grating, flat field, dou-
ble exit monochromator equipped with a photomultiplier detector
(Hamamatsu R3896) and a Princeton Instruments PIMAX II gated
intensified CCD camera, using a RB Gen II intensifier, a ST133 con-
troller and a PTG pulser. Signals from the photomultiplier (kinetic
traces) were processed by means of a LeCroy 9360 (600 MHz,
5Gs/s) digital oscilloscope, while transient difference spectra were
recorded with the CCD camera in a single beam configuration,
acquiring first the light transmitted by the sample ground state
and then after laser excitation at various delays and gate widths.
A pulse energy below 10 mJ/pulse was used in all measurements.
mation of hydroxyl radicals during photoexcitation of Na W10O32
4
in water solutions. In particular, we focus our attention on the
formation mechanism of these intermediates and on their involve-
ment in the oxidation of propan-2-ol through EPR spin trapping
and laser flash photolysis experiments.
The ability to control the chemoselectivity of the propan-2-ol
oxidation process is another objective of this work. With this in
mind, we also investigated the possible effect of a microporous
siliceous matrix to exert a specific substrate recognition for improv-
ing the selectivity of the process. In fact, recent developments in the
field of photocatalysis indicate that heterogenisation of Na W10O32
4
ꢀ
with solid supports is a suitable means in order to provide selec-
tive and recyclable photocatalytic systems [2,13–16]. In particular,
it has been shown that decatungstate entrapped inside polymeric
membranes induces the partial oxidation of several water soluble
alcohols with accumulation in solution of carbonylic products up
to a substrate conversion of 30% without appreciable evidence of
CO2 [22]. Moreover, we reported recently that photoexcitation of
Na W10O32 heterogenized with microporous silica induces glyc-
erol oxidation mainly to glyceraldehyde and dihydroxyacetone
with only negligible amounts of CO2 in spite of the observed for-
Samples were purged 20 with argon when requested.
2.4. Photocatalytic experiments
Photocatalytic experiments were carried out inside
a
−
4
closed Pyrex tube at 298 ± 1 K. Na W10O32 (2 × 10 M) or
4
Na W
O /SiO (8 g/L) was placed in an aqueous solution (3 mL)
4
10 32 2
−
1
4
containing propan-2-ol (10 M). The photoreactor was joined
to a balloon filled with O2 and irradiated by an external Helios
Q400 Italquartz medium-pressure Hg lamp, selecting wave-
lengths higher than 300 nm with a cut off filter (photon flux was
•
mation of OH radicals [23].
−
2
0
.075 W cm ). Then, samples were analyzed by a HP 6890 gas
2
. Experimental
chromatograph equipped with a FID and with a HPWAX capillary
column. Quantitative analysis for acetone was performed with
calibration curve obtained with authentic sample. Each exper-
iment was repeated three times in order to evaluate the error,
which remained in the ± 5% interval around mean values. Control
2.1. Photocatalyst preparation and characterization
Reagents and solvents were purchased from Sigma in the
highest purities available and used without further purification.
Sodium decatungstate (Na W10O32) was synthesized following
a literature procedure [24]. The heterogeneous photocatalyst
experiments were carried out by irradiating SiO suspended in
2
−
1
the solution containing propan-2-ol (10 M) or keeping the
photocatalyst dispersed in the solution in the dark. The amount
of CO2 eventually formed was measured following a procedure
described in detail in Ref. 23 that uses a pH meter BasiC 20 CRISON
equipped with a gas sensing probe (Crison 9666). The yield of CO2
4
(
Na W10O32/SiO ) was prepared by hydrolysis of tetraethy-
4 2
lorthosilicate (TEOS) in the presence of an acid aqueous solution
of Na W10O32, as recently published in detail for the entrapment
4

A. Molinari et al. / Journal of Molecular Catalysis A: Chemical 372 (2013) 23–28
25
(
b)
(
a)
6
6,0x10
(c)
(
b)
6
3,0x10
(
a)
0,0
400
500
600
3425
3450
3475
3500
wavelength (nm)
magnetic field (gauss)
Fig. 2. Changes in emission intensity measured at 455 nm of: (a) aqueous solution
−
4
−4
of Na4W10O32 (2 × 10 M) and coumarin (1 × 10 M) before irradiation, (b) sample
Fig. 1. EPR spectra obtained after 30 s irradiation (ꢀ > 300 nm) of aqueous solu-
ꢀ
ꢀ
−2
−3
(a) after 30 irradiation (ꢀ > 300 nm), (c) sample (a) irradiated (30 , ꢀ > 300 nm) but
tions of DMPO (3 × 10 M) containing: (a) Na4W10O32 (2 × 10 M), (b) powder
in the presence of catalase.
dispersions of Na4W10O32/SiO2.
−
was referred to the number of carbon atoms present in the alcohol
molecule.
and O2 , which presents a typical 12-lines EPR spectrum [30,31].
Literature data indicate that a maximum level of approximately
•
•
Other photochemical experiments have been carried out irradi-
3% [DMPO–OH] should originate from [DMPO–OOH] . Consider-
ing that the spectrum of Fig. 1 (curve a) is the only observed during
irradiation, we conclude that, although in our photocatalytic sys-
ꢀ
ating (ꢀ > 300 nm, 30 ) aerated aqueous solutions (3 mL) containing
−
4
−4
Na W10O32 (2 × 10 M) and coumarin (Sigma, 1 × 10 M).
4
5
−
tem reductive activation of O by W10O32 can lead to superoxide
2
After irradiation the fluorescence spectrum (ꢀexc = 332 nm)
•
•
of 7-hydroxycoumarin, eventually formed, was recorded
species, [DMPO–OOH] is a negligible source of [DMPO–OH] . Sec-
ond, it has been demonstrated that in an oxidant environment
(
ꢀemiss = 455 nm) at room temperature with a Jobin Yvon Spex
Fluoromax II spectrofluorimeter equipped with a Hamamatsu
R3896 photomultiplier. Both emission and excitation slits were set
at 5.0 nm during the measurements.
(such as photoexcited aqueous suspensions of TiO ) DMPO can
2
be directly oxidized to the corresponding radical cation. Its subse-
•
quent reaction with water leads to the formation of [DMPO–OH]
32]. In order to confirm that in our case Eq. (1) is the source
[
•
•
3
. Results and discussion
of OH radicals and that the detection of [DMPO–OH] is not the
result of other redox processes involving the employed spin trap,
we decided to use coumarin as fluorescent probe. In fact, it is
reported that coumarin produces 7-hydroxycoumarin, a strongly
luminescent compound, by reaction with hydroxyl radicals accord-
ing to Eq. (3). This method has been already successfully applied
for the detection of hydroxyl radicals generated by photoexcited
•
3
.1. OH radicals formation
EPR spin-trapping investigation is a powerful technique for
detecting the formation of short-lived radicals and it has been fruit-
fully employed in photochemical studies on polyoxometalates in
order to better understand primary processes occurring after the
excitation [25–28]. This technique is based on the ability of some
molecules such as nitrones to trap short-lived radicals to give para-
magnetic nitroxides stable enough to be successfully detected and
studied [29]. The nature of the trapped radical can be often identi-
fied by the parameters obtainable from the EPR spectrum.
TiO [33]. Therefore, some experiments have been carried out irra-
2
−
4
diating aqueous solutions containing Na W10O32 (2 × 10 M) and
4
−
4
coumarin (1 × 10 M). The comparison between emission spectra
before and after irradiation (Fig. 2 curves (a) and (b)) points out the
formation of the fluorescent 7-hydroxycoumarin.
−
3
O
Aqueous solutions of Na W10O32 (2 × 10 M) have been irra-
O
HO
O
4
O
diated inside the EPR cavity (ꢀ > 300 nm) in the presence of DMPO
OH.
+
−
2
(
3 × 10 M) as spin trap. Few seconds photoirradiation causes the
(3)
prompt formation of a quartet (1:2:2:1, aN = aH = 14.6 G), which is
shown in Fig. 1 (curve a). Signal pattern and coupling constant val-
•
3.2. OH radicals origin
•
ues are in agreement with the trapping of OH radicals to form the
•
paramagnetic adduct [DMPO–OH] according to Eq. (2) [25]. Con-
Laser flash photolysis experiments have been carried out on
deaerated solutions of Na W10O in the presence of DMPO. The
absorbance growth at 780 nm after a 7 ns laser pulse is in agree-
ment with the formation of the transient wO (Fig. 3 curve (a))
trol experiments show that no signal is observed neither in the dark
nor during irradiation but in the absence of decatungstate.
4
32
H
.
+
OH
H
OH
[
9]. Moreover, transient absorbance changes in the wavelength
range 450–900 nm in the insert of Fig. 3 show the formation of the
N
N
O
5−
one-electron-reduced form of the decatungstate W10O32 , which
absorbs at 780 nm too [3]. It is also seen that this species survives
on a much longer time scale. The above results are a clear indica-
tion that photochemical excitation induces W10O32 reduction to
W10O32 , so confirming that Eq. (1) does occur. Curve (b) of Fig. 3
O
(2)
This result suggests that reaction (1) is occurring in our sys-
tem. However, some other sources of the quartet relative to the
4−
•
5−
paramagnetic [DMPO–OH] species should be ruled out. First, it has
•
5−
been previously reported that the [DMPO–OH] species may origi-
nate from degradation of the adduct [DMPO–OOH] between DMPO
shows that W10O32 is not accumulated in detectable amounts in
•
the absence of DMPO, likely because the photogenerated hydroxyl

2
6
A. Molinari et al. / Journal of Molecular Catalysis A: Chemical 372 (2013) 23–28
0
0
0
,08
,04
,00
0
0
0
,04
,02
,00
450
600
750
900
wavelength (nm)
(a)
(
b)
0
25
50
75
100
125
Time (s)
2
00
400
600
•
Fig. 5. Curve (a): fixed-field EPR signal intensity of the [DMPO–OH] adduct in
time upon irradiation (ꢀ > 300 nm) of an aqueous solution containing Na4W10O32
Time (ns)
−
3
−2
(
3 × 10 M) and DMPO (3 × 10 M); curve (b): the same experiment but in the
−1
presence of propan-2-ol (10 M).
Fig. 3. Kinetics profiles observed at 780 nm obtained after laser excitation at 355 nm
−4
of a deaerated aqueous solution of Na4W10O32 (2 × 10 M). Curve (a): in the pres-
−3
ence of DMPO (3 × 10 M). Curve (b): in the absence of DMPO. Insert: corresponding
•
seen that the signal intensity of [DMPO–OH] is significantly lower,
but not negligible, when irradiation is carried out in the presence
of catalase. Curve (c) shows the EPR signal intensity when the pre-
vious experiment with catalase is carried out also in the absence of
O . It is seen that the direct H O oxidation leads to the formation of
detectable amounts of [DMPO–OH] just after few seconds irradia-
tion. Later, when most of the decatungstate has been reduced and
can not be reoxidized due to the absence of O , the [DMPO–OH]
transient absorbance spectrum.
radicals are able to quickly reoxidize W10O325 if they are not scav-
enged by the spin trap.
The decatungstate in its phoreduced form is able to initiate
reductive activation processes of O2 leading up to hydrogen per-
oxide. Therefore, the one-electron reduction of H O according to
Eq. (4) should be considered as an additional reaction pathway for
the formation of hydroxyl radicals. Detection of OH radicals during
photolysis of acetonitrile solutions of decatungstate in the presence
of an organic substrate has been previously explained on this basis
−
2
2
•
•
2
2
2
signal decreases quickly.
•
In line with the statement that both Eqs. (1) and (3) contribute
to the formation of OH radicals, Fig. 2 (curve c) shows that a lesser
luminescence by 7-hydroxycoumarin is observed when aqueous
solutions of Na W10O32 and coumarin are irradiated as described
•
[
25].
4
in Section 3.1 but in the presence of catalase. This can be due to
H O + W10O325 → W10O32 + OH• + OH−
−
4−
(4)
2
2
the fact that a lower amount of H O2 can be accumulated and,
2
consequently, lesser amounts of hydroxyl radicals are formed.
Some experiments carried out in the presence of catalase enable
us to establish the origin of hydroxyl radicals during irradiation of
water solution of Na W10O32. This enzyme is able to catalyze H O2
4
2
3.3. Photoexcitation in the presence of propan-2-ol
disproportionation with very high efficiency, so avoiding its accu-
mulation. On this basis, we can assume that the contribution of
Eq. (4) to the production of hydroxyl radicals is negligible when
photoirradiation is carried out in the presence of catalase. Fig. 4
shows the EPR signal intensity changes of the [DMPO–OH] adduct
in the absence (curve a) and in the presence of catalase (curve b).
A comparison between these curves allows us to conclude that,
Additional spin trapping investigations have been carried out on
−
1
aqueous solutions of Na W10O32 containing propan-2-ol (10 M).
4
•
The only detected EPR signal is the same quartet of the [DMPO–OH]
•
adduct (Fig. 1), indicating that Eq. (1) occurs also in the presence of
alcohol. However, Fig. 5 shows that the intensity of this adduct is
significantly lower than that observed in pure water. This behavior
can be ascribed to two possible reactions involving the alcohol. In
fact, propan-2-ol can compete with water for the reaction with wO
the hydroxyl radicals formed during photoexcitation of Na W10O32
4
originate from both H O oxidation and H O reduction. In fact, it is
2
2
2
•
according to Eq. (5) and, with DMPO for the reaction with OH radi-
•
cals (Eq. (6)). It is expected that the reaction between OH radicals
and propan-2-ol leads to the formation of water and hydroxy alkyl
radicals.
(
a)
wO + (CH ) CHOH → (CH ) C OH + W10O32 + H+
•
5−
(5)
(6)
3
2
3 2
•
•
OH + (CH ) CHOH → (CH ) C OH + H O
3
2
3 2
2
(
b)
Laser flash photolysis experiments have been performed on
deaerated aqueous solutions of Na W10O32 in the presence of
4
propan-2-ol. The kinetics profile is in line with the formation of
the mono-reduced form of decatungstate (Fig. 6). This is further
confirmed by the transient absorbance changes in the wavelength
range 450–900 nm reported in the insert of Fig. 6. It is also seen an
absorption at about 630 nm which can be ascribed to the formation
(
c)
0
25
50
75
100
125
Time (s)
6−
of W10O32 [3]. This result is in agreement with previous observa-
tions indicating that a further electron transfer from hydroxy alkyl
•
Fig. 4. Fixed-field EPR signal intensity of the [DMPO–OH] adduct vs. time upon
−
3
irradiation of: (a) aerated aqueous solution of Na4W10O32 (2 × 10 M) and DMPO
5−
radicals to W10O32 can generate the two electron reduced poly-
−
2
(
3 × 10 M); (b) as in sample (a) but in the presence of catalase; (c) deaerated aque-
−3 −2
oxoanion and the carbonyl derivative according to Eq. (7) [34,35].
This very fast reaction gives reason of the lack of any EPR signal due
ous solution of Na4W10O32 (2 × 10 M) and DMPO (3 × 10 M) in the presence of
catalase.

A. Molinari et al. / Journal of Molecular Catalysis A: Chemical 372 (2013) 23–28
27
4
.0E-03
CO2
acetone
CO2
0
0
0
0
,03
,02
,01
,00
3.0E-03
acetone
0
,12
2.0E-03
0,08
0,04
1.0E-03
0,00
4
50
600
750
900
wavelength (nm)
0.0E+00
1
50
300
450
600
0
20
40
60
80
Time (ns)
Irradiation time (min)
Fig. 6. Kinetics profile observed at 780 nm obtained after excitation at 355 nm of a
Fig. 7. Acetone and CO2 concentrations vs. irradiation time obtained irradiating
−
4
−1
deaerated aqueous solution of Na4W10O32 (2 × 10 M) and propan-2-ol (10 M).
−4
(
ꢀ > 290 nm) Na4W10O32 (2 × 10 M, circles) and Na4W10O32/SiO2 (8 g/L, triangles)
Insert: corresponding transient absorbance spectrum.
in aqueous solutions containing propan-2-ol 1 M.
to the adduct between DMPO and hydroxy alkyl radicals during the
EPR spin trapping experiments.
homogeneous Na W10O32, photoexcited Na W10O32/SiO2 accu-
4
4
mulates primary oxidation products with negligible amounts of
over-oxidation products. On this basis, we were prompted here
to evaluate the possibility to increase the chemoselectivity of
•
5−
6−
(
CH ) C OH + W10O32 → (CH ) CO + W10O32 + H+
(7)
3
2
3 2
Gas chromatographic analyses indicate that acetone concen-
tration increases quite linearly during the first thirty minutes
Na W10O32 in the photocatalytic oxidation of propan-2-ol by
4
using Na W10O32/SiO2.
4
−
4
irradiation of aerated aqueous solutions of Na W10O (2 × 10 M)
4
32
EPR spin-trapping investigation indicate that heterogenization
−1
and propan-2-ol (10 M) (see Supplementary data). It is also
observed that, when the experiment is performed in the presence
of catalase, which reduces the amount of OH radicals, the ace-
does not affect the ability of photoexcited Na W10O32 to produce
4
hydroxyl radicals. In fact, photoexcitation of aqueous suspensions
•
of Na W10O32/SiO in the presence of DMPO leads to the formation
4
2
tone yield decreases of about 60% This result allows us to conclude
•
of the [DMPO–OH] adduct (Fig. 1, curve b), in agreement with the
results obtained in homogeneous solution.
•
that OH radicals are involved in the photocatalytic oxidation of
propan-2-ol.
The results concerning the photocatalytic activity of
CO2 deriving from over-oxidation of propan-2-ol must be
counted among the possible reaction products, since hydroxyl radi-
cals are known to be very powerful reacting species that are able to
overoxidize various organic substrates. Indeed, the results reported
in the first row of Table 1 shows that the photocatalytic oxidation
Na W
O
32/SiO2 (8 g/L) in aqueous solutions containing propan-
4
10
−1
2
-ol (10 M) are shown in the second row of Table 1. Propan-2-ol
is oxidized to acetone and CO . A comparison between the
acetone/CO ratios of rows 1 and 2 allows us to conclude that the
solid support inhibits in some extent the over-oxidation to CO2.
2
2
of propan-2-ol by Na W10O yields considerable amounts of CO2.
4
32
A main advantage of heterogenization is that Na W10O32/SiO2
4
This result may be of interest for the development of photocat-
alytic systems for pollutants abatement but, on the other hand, it is
a serious problem for possible application in synthesis. Efficiency
of a photocatalyst is usually evaluated also in terms of its stability.
The system here investigated presents a serious inconvenient from
this point of view. In fact, spectroscopic evidences indicate that at
is a robust photocatalyst. In fact, the same photocatalyst has been
used for repeated experiments after simple washing and drying
at 70 C for 10 min. The decrease in activity, in terms of detected
end products, is less than 10% after three repeated cycles. More-
over, control experiments show that any kind of photoactivation
of the silica matrix can be excluded since irradiation of a disper-
◦
neutral pH Na W10O undergoes quite fast degradation to smaller
4
32
sion of SiO does not lead to any oxidation product. Heterogenized
2
fragments that do not absorb light at ꢀ > 300 nm.
decatungstate does not oxidize propan-2-ol in the absence of light.
Finally, UV–vis spectra of solutions recovered after the irradiation
3
.4. Photocatalytic properties of heterogenized Na W10O32
of Na W10O32/SiO2 show that decatungstate is not released from
4
4
the support.
Heterogeneous systems represent a suitable means to tune
A significant increase of the matrix effect on the selectiv-
ity is observed when the photocatalytic experiments are carried
out in the presence of propan-2-ol ten times more concentrated
(1 M). Fig. 7 shows the yields of acetone and CO2 as a function of
efficiency and selectivity of photocatalytic processes through the
control of the microscopic environment surrounding the photoac-
tive center [2,13–16,22]. In particular, it has been already evidenced
that silica surface is able to discriminate among reactants, favoring
alcohol adsorption in close proximity to the photoactive species
irradiation time for both homogeneous Na W10O32 (circles) and
4
Na W10O32/SiO2 (triangles). It is seen that CO2 is the main reac-
4
[
16,23,36,37]. As a consequence, partially oxidized products can
tion product in the homogeneous experiment, while, its formation
accumulate in the solution bulk without over-oxidation to carbon
is quite negligible when Na W10O32/SiO2 is used. In particular,
4
dioxide. We reported recently about the entrapment of Na W10O32
acetone/CO2 ratios, obtained at similar propan-2-ol conversion,
4
in a silica matrix by a sol–gel procedure [23]. This material
are 7.5 with Na W10O32/SiO2 and only 0.7 with homogeneous
4
(
Na W10O32/SiO ), that contains 30% (w/w) of decatungstate and
decatungstate. This very important result can be ascribed to the
known ability of silica of promoting alcohol adsorption, allowing
the enrichment of substrate on the surface and favoring the detach-
ment of acetone, which prevents its over-oxidation [23,36,37].
4
2
that is characterized by micropores (7 A˚ and 13 A˚ ) and mesopores
30 A˚ ), strongly affects the selectivity of glycerol oxidation
process. In particular, contrary to what obtained with
(

2
8
A. Molinari et al. / Journal of Molecular Catalysis A: Chemical 372 (2013) 23–28
Table 1
Photocatalytic properties of Na4W10O32 and of Na4W10O32/SiO2 in the oxidation of propan-2-ol.
a
Run
Photocatalytic system
Acetone (mol/L)
CO2 (mol/L)
Molar ratio acetone/CO2
−
−
3
4
−3
−4
1
2
Na4W10O32 propan-2-ol 0.1 M
Na4W10O32/SiO2propan-2-ol 0.1 M
1.42 × 10
6.47 × 10
1.74 × 10
0.82
1.41
4.60 × 10
a
−4
In a typical experiment, Na4W10O32 (2 × 10 M) or Na4W10O32/SiO2 (8 g/L) was suspended in an aqueous solution containing propan-2-ol and irradiated (30 min,
> 290 nm) at 298 ± 1 K and 760 Torr of O2. Reported values are the mean of three repeated experiments with ± 5% of precision. The chosen amount of photocatalysts warrants
ꢀ
the absorption of 90% of the incident radiation at 313 nm, which is the emission line of the employed light source closest to the absorption maximum of decatungstate at
3
23 nm.
4
. Conclusions
[6] M.D. Tzirakis, I.N. Likakis, M. Orfanopoulos, Chem. Soc. Rev. 38 (2009)
609–2621.
2
[
[
7] A. Maldotti, A. Molinari, R. Amadelli, Chem. Rev. 102 (2002) 3811–3836.
8] C. Tanielian, K. Duffy, A. Jones, J. Phys. Chem. B 101 (1997) 4276–4282.
This study demonstrates that hydroxyl radicals can be formed
irradiating aqueous solutions of Na W10O32. We show that these
[9] D.C. Duncan, M.A. Fox, J. Phys. Chem. A 102 (1998) 4559–4567.
10] C.L. Hill, Z. Zheng, Chem. Commun. (1998) 2467–2468.
11] A. Hiskia, E. Papaconstantinou, Inorg. Chem. 31 (1992) 163–167.
12] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037–3058.
4
[
[
[
species originate from both H O oxidation and H O reduction and
2
2
2
that they cause propan-2-ol oxidation.
Photocatalytic investigation indicates that Na W10O32 is able to
[13] M.D. Tzirakis, I.N. Lykakis, G.D. Panagiotou, K. Bourikas, A. Lycourghiotis, C.
Kordulis, M. Orfanopoulos, J. Catal. 252 (2007) 178–189.
4
oxidize propan-2-ol up to CO with high yield. This result may be
2
[
14] A. Molinari, A. Bratovcic, G. Magnacca, A. Maldotti, Dalton Trans. 39 (2010)
826–7833.
[15] S. Farhadi, Z. Momeni, J. Mol. Catal. A: Chem. 277 (2007) 47–52.
of interest for the development of photocatalytic systems aimed to
pollutants degradation. On the other hand, we have evidence that
the decatungstate dissolved in water is not stable and undergoes
complete degradation in few hours.
7
[
[
16] A. Maldotti, A. Molinari, F. Bigi, J. Catal. 253 (2008) 312–317.
17] A. Mylonas, A. Hiskia, E. Androulaki, D. Dimotikali, E. Papaconstantinou, Phys.
Chem. Chem. Phys. 1 (1999) 437–440.
The strong oxidative powerful of Na W10O32 is a drawback in
[18] P. Kormali, T. Triantis, D. Dimotikali, A. Hiskia, E. Papaconstantinou, Appl. Catal.
B: Environ. 68 (2006) 139–146.
4
photocatalysis when the main purpose is the selective accumula-
[
[
19] I. Texier, J.A. Delaire, C. Giannotti, Phys. Chem. Chem. Phys. 2 (2000) 1205–1212.
20] R.R. Ozer, J.L. Ferry, J. Phys. Chem. B 104 (2000) 9444–9448.
tion of valuable reaction intermediates. Entrapment of Na W10O32
4
into a microporous silica matrix gives a robust photocatalyst where
the decatungstate structure is preserved. This material is able to
catalyze the photooxidation of propan-2-ol to acetone with appre-
ciable chemoselectivity. In particular, optimal conditions have been
found allowing the solid photocatalyst to yield acetone as main
product inhibiting over-oxidation processes that lead to carbon
dioxide.
[21] T. Yamase, Inorg. Chim. Acta 76 (1983) L25–L27.
[22] M. Bonchio, M. Carraro, G. Scorrano, E. Fontananova, E. Drioli, Adv. Synth. Catal.
345 (2003) 1119–1126.
[
23] A. Molinari, A. Maldotti, A. Bratovcic, G. Magnacca, Catal. Today (2011),
http://dx.doi.org/10.1016/j.cattod.2011.11.033.
[24] F. Bigi, A. Corradini, C. Quarantelli, G. Sartori, J. Catal. 250 (2007)
22–230.
25] A. Maldotti, R. Amadelli, V. Carassiti, A. Molinari, Inorg. Chim. Acta 256 (1997)
09–312.
26] D. Dondi, M. Fagnoni, A. Molinari, A. Maldotti, A. Albini, Chem. Eur. J. 10 (2004)
42–148.
27] A. Molinari, M. Montoncello, H. Rezala, A. Maldotti, Photochem. Photobiol. Sci.
(2009) 613–619.
2
[
[
[
3
1
Appendix A. Supplementary data
8
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.molcata.2013.01.037.
[28] D. Dondi, M. Ravelli, M. Fagnoni, A. Mella, A. Molinari, A. Maldotti, A. Albini,
Chem. Eur. J. 15 (2009) 7949–7957.
[
[
[
[
29] E.G. Janzen, Acc. Chem. Res. 4 (1971) 31–40.
30] E. Finkelstein, G.M. Rose, E.J. Rauckman, Mol. Pharmacol. 21 (1982) 262–265.
31] G.R. Buettner, Free Radic. Res. Commun. 19 (1993) S79–S87.
32] Y. Nosaka, S. Komori, K. Yawata, T. Hirakawa, A.Y. Nosaka, Phys. Chem. Chem.
Phys. 5 (2003) 4731–4735.
References
[
33] H. Czili, A. Horvath, Appl. Catal. B: Environ. 81 (2008) 295–302.
[
[
[
[
[
1] A. Hiskia, A. Mylonas, E. Papaconstantinou, Chem. Soc. Rev. 30 (2001) 62–69.
2] A. Maldotti, A. Molinari, Topics Curr. Chem. 303 (2011) 185–216.
3] C. Tanielian, Coord. Chem. Rev. 178 (1998) 1165–1181.
4] M. Fagnoni, D. Dondi, D. Ravelli, A. Albini, Chem. Rev. 107 (2007) 2725–2756.
5] G. Palmisano, V. Augugliaro, M. Pagliaro, L. Palmisano, Chem. Commun. (2007)
[34] M.A. Fox, R. Cardona, E. Gaillard, J. Am. Chem. Soc. 109 (1987) 6347–6354.
[35] E. Papaconstantinou, Chem. Soc. Rev. 18 (1989) 1–31.
[36] A. Molinari, R. Amadelli, A. Mazzacani, G. Sartori, A. Maldotti, Langmuir 18
(2002) 5400–5405.
[37] A. Molinari, A. Maldotti, A. Bratovcic, G. Magnacca, Catal. Today 161 (2011)
64–69.
3
425–3437.