Selective solar photocatalytic oxidation of ethylbenzene on C, N, and S doped TiO2

Article abstract of DOI:10.1007/s11164-011-0298-7

The liquid-phase photo-oxidation of ethylbenzene (EB) is investigated in solar light with air/O₂/N₂ at atmospheric pressure, in a batch reactor using acetonitrile medium. It is carried out over TiO₂ doped with C, N, and S

Full text of DOI:10.1007/s11164-011-0298-7

Res Chem Intermed (2011) 37:901–918
DOI 10.1007/s11164-011-0298-7
Selective solar photocatalytic oxidation of ethylbenzene
on C, N, and S doped TiO2
Basavaraju Srinivas
Mucharla Rajesh Valluri Durga Kumari
Machiraju Subrahmanyam
•
Police Anil Kumar Reddy
•
•
•
•
Bhudev Ranjan De
Received: 3 January 2011 / Accepted: 15 February 2011 / Published online: 9 March 2011
Ó Springer Science+Business Media B.V. 2011
Abstract The liquid-phase photo-oxidation of ethylbenzene (EB) is investigated
in solar light with air/O /N at atmospheric pressure, in a batch reactor using ace-
2
2
tonitrile medium. It is carried out over TiO doped with C, N, and S (TCNS)
2
photocatalyst samples. The photocatalytic oxidation yielded acetophenone (33%)
and 1-phenylethanol (21%) at 56% conversion of EB during a 6-h irradiation time.
This product distribution indicates that C–H bond activation has occurred only at the
alkyl chain. The effects of the EB to water content ratio, amount of photocatalyst
and its sustainability, pH, have been studied. It is illustrated that the reaction carried
out by this environmentally friendly photocatalysis is truly heterogeneous under
mild conditions using solar light and no waste generation. An optimum loading of
TCNS5 was observed for the photo-oxidation of EB.
Keywords Ethylbenzene oxidation ꢀ Acetophenone ꢀ Titanium dioxide ꢀ
Photo-oxidation ꢀ C, N, and S doping
Introduction
The oxidation of benzyl C–H bonds constitutes one of the most fundamental
transformations in organic synthesis. The aromatic ketones which are important
chemical intermediates for the synthesis of various perfumes, drugs, and pharma-
ceuticals production are made by Frieda–Crafts acylation of aromatics by acyl
halide or acid anhydride, using stoichiometric quantities of homogeneous acid. This
B. Srinivas ꢀ P. Anil Kumar Reddy ꢀ M. Rajesh ꢀ V. Durga Kumari ꢀ M. Subrahmanyam (&)
Inorganic and Physical Chemistry Division, Indian Institute of Chemical Technology, Hyderabad
500607, India
e-mail: subrahmanyam@iict.res.in
B. R. De
Department of Chemistry and Chemical Technology, Vidyasagar University, Midnapur, India
123

902
B. Srinivas et al.
leads to the formation of a large volume of toxic and corrosive wastes. Previously,
efforts have been made to produce benzylic ketones by oxidizing the methylene
group of alkyl aromatics using stoichiometric quantities of KMnO as an oxidizing
4
agent. However, in the above stoichiometric oxidation, the volume of waste
produced is also very large, and separating reactants and products from the liquid
reaction mixture is difficult.
In this context, the utilization of ethylbenzene (EB) in the xylene stream of the
petrochemical industry to more value-added products is a worthwhile approach.
In fact, the oxidation of EB has not drawn much attention due to the complexity of
the products that could be formed. The reaction conditions are harsh, the product
selectivity is poor, often corrosive promoters like bromide anions are used along
with the catalyst, the separation of catalyst from the reaction mixture is difficult, the
catalyst cannot be reused and also a lot of tarry waste is produced. It is, therefore,
of great practical interest to develop a more efficient, easily separable, reusable, and
environmentally friendly catalyst for this reaction. Processes based on oxidation,
particularly replacing stoichiometric oxidation such as KMnO , Ag O, and CrO , by
4
2
3
photocatalysis using solar energy at room temperature is, self-evidently, a sensible
strategy to develop [1–6].
Photocatalytic organic synthesis arises from synthetic, mechanistic, or environ-
mental application that represents a key strategy for the development of sustainable
methods for chemical transformations [7, 8]. Reviews on the photocatalytic
oxidation of various hydrocarbons are reported in the literature [9, 10]. Among the
oxidation catalyst requirements, photoexcited semiconductors such as TiO has
2
attracted much attention in the last two decades. Semiconductor-mediated
photocatalytic oxidation offers a facile and cheap method. Among various oxide
semiconductor photocatalysts, TiO has proved to be the most suitable catalyst for
2
widespread environmental applications because of its biological and chemical
inertness, strong oxidizing power, non-toxicity, and long-term stability against
photo and chemical corrosion. However, the most negative property of TiO is the
2
need for using an UV wavelength of \387 nm (availability in solar light is \5%),
as the excitation source. In this context, we explored here the possibility to use TiO2
anion-modified material in order to harvest a greater portion of solar radiation.
Several works reported that doping TiO with anions such as carbon, nitrogen,
2
sulphur, boron, and fluorine shifts the optical absorption edge of TiO towards lower
2
energy, thereby increasing the photocatalytic activity in the visible light region.
Furthermore, DTA analysis studies reported on TiO using thiourea as a precursor
2
illustrates the role of dopant on the calcination temperature [11–14]. In the present
investigation, the main focus is on shifting the absorption edge of TiO to the visible
2
light region by introducing C, N, and S into the TiO lattice structure. The present
2
results obtained provide a simple route for the preparation of C, N, and S doped
TiO with enhanced photocatalytic activity compared to bare TiO under visible
2
2
light irradiation for EB oxidation to acetophenone. To the best of our knowledge, no
report is available in the literature on the photocatalytic oxidation of EB except for
two, wherein in one report, environmental impact, high cost, and their hydrothermal
and chemical stability are questionable issues [15] and the other one deals with UV-
active TiO alone with meager conversion and yields [16]. Based on these reports
2
1
23

Selective solar photocatalytic oxidation of ethylbenzene
903
and building on our previous expertise [7], we report herein the selective activation
of benzylic C–H bonds of EB (difficult to oxidize due to the absence of an electron-
donating aromatic ring-activation group) to acetophenone and 1-phenylethanol
selectively using solar energy (Scheme 1).
The data obtained after 6 h of irradiation time and other optimization studies
needed for any developmental work are reported in this paper.
Experimental
Materials and methods
All of the chemicals in the present work were of analytical grade and used as such
without further purification. EB was of Fluka grade, titanium isopropoxide was from
Sigma-Aldrich Chemie Munich, Germany, and acetonitrile (HPLC grade) was
obtained from Ranbaxy Limited, India. All of the experiments were carried out with
acetonitrile as the medium.
Preparation and characterization of catalyst
The catalyst samples for EB oxidation were C, N, and S doped TiO prepared,
2
characterized, and reported in our earlier paper using titanium isopropoxide as the
precursor for titanium and thiourea as the source for carbon, nitrogen, and sulphur
elements [11]. The weight (%) of thiourea doped TiO was controlled at 0, 1, 3, 5,
2
1
0, 15, 25, and 40 wt% and the samples obtained were labeled as TCNS0, TCNS1,
TCNS3, TCNS5, TCNS10, TCNS15, TCNS25, and TCNS40, respectively. All of
the catalysts were characterized by techniques such XRD, XPS, and FTIR, SEM,
BET surface area and UV–Vis DRS, CHNS elemental analysis [11].
Photocatalytic experiments
The substrate EB was taken in acetonitrile solvent. All of the photocatalytic
experiments were carried out at the same concentrations until unless otherwise
stated. Prior to illumination of the suspended solution, dark (adsorption) experi-
ments were carried out for better adsorption of the EB on the catalyst. For solar
H
H3C
O
3
H C
OH
CH2CH3
Solar Light
Air (1atm)
Temp = 35°C
CH CN
3
Catalyst
Ethylbenzene
Acetophenone
1-Phenylethanol
Scheme 1 EB oxidation over photocatalyst using solar irradiation
123

904
B. Srinivas et al.
-
2
experiments, the EB solution of 2.73 9 10 M and the calculated amount of water
was taken in a Borosil glass tube with a known amount of the catalyst sealed with
Aldrich rubber septum. The solar photocatalytic experiments are conducted in a
reactor wherein water was circulated around the reaction chamber, thus, maintaining
the reaction at a constant temperature. The solar experiments were carried out
during the time period 10.00 a.m. to 3.00 p.m. in May and June 2010 at Hyderabad,
India.
Analysis
The EB oxidation was monitored by a Shimadzu SPD-20A HPLC system using a
C-18 Phenomenex reverse-phase column with acetonitrile/water (70/30 v/v %) as
-
1
the mobile phase at a flow rate of 1 mL min . The samples were collected at
regular intervals and filtered through Millipore microsyringe filters (0.2 lm).
Results and discussion
Characterization
BET surface area
The surface area of TCNS catalysts calcined at 400 °C is shown in Table 1. The
high surface area of the prepared catalysts is due to the nano size of the particles. It
is also observed that the surface area of the catalysts increases with the increase in
the ratio of thiourea to TiO . This can be attributed to the decreasing of the particle
2
sizes, as reported in the XRD analysis of the samples [11].
FTIR spectra
Figure 1 shows the FTIR spectra of TCNS0 and TCNS5 catalysts calcined at
1
-
4
the stretching vibration and bending vibration of the hydroxyl group respectively
00 °C. The absorption bands 2,800–3,500 and 1,600–1,680 cm are assigned to
present on the surface of TiO catalyst [17, 18]. The presence of surface hydroxyl
2
-
1
groups are substantiated by XPS of O1s spectra [11]. The band around 1,730 cm
Table 1 BET surface area and
particle size of the TCNS
catalysts
Catalyst
Particle size
by XRD (nm)
BET surface
2
area (m /g)
TCNS0
TCNS1
TCNS3
TCNS5
TCNS10
TCNS15
5.8
5.2
5.1
4.6
4.2
3.8
80.53
82.47
83.97
89.14
92.98
124.28
1
23

Selective solar photocatalytic oxidation of ethylbenzene
905
-
1
is attributed to the carbonyl group, and the bands at 1,130 and 1,040 cm
correspond to the nitrite and hyponitrite groups present in TCNS5 and they are
absent in TCNS0, which shows the successful doping of nitrogen into the lattice of
?
4
TiO [11, 19, 20]. It is evident from the spectra that no peak corresponds to the NH
2
-
1
ion (absence of 3,189 and 1,400 cm ) in the prepared TCNS catalysts, which
shows that N is present only in the form of nitrite and hyponitrite species [18].
SEM
The surface morphology of the TCNS photocatalyst is studied by SEM and the
micrographs are presented in Fig. 2. The samples appear to be smaller particles. The
SEM images of the TCNS0 (bare) and CNS-doped (TCNS5) catalysts showed that
the particle morphology seems spherical in both of the images Fig. 2. The
photograph of the TCNS5 sample exhibits well-dispersed particles and the particle
is homogeneous with the formation of fine and well-dispersed particles. The
investigation made for the sample used after the reaction (Fig. 2c) is illustrated later
in recyclable activity studies. CHNS chemical analysis confirms that the amount of
thiourea doped corresponds to 1% C, N, and S, which was observed for TCNS5. The
higher activity on optimized reaction conditions of TCNS5, for the pre- and the
post-reaction samples of TCNS5 chemical analysis showed 0.80–0.95% C, N, and
S. This indicates that there were no surface changes on the catalysts after usage.
UV–Vis DRS
The UV–Vis diffuse reflectance spectra (DRS) of TCNS catalysts are shown in
Fig. 3. It is seen from Fig. 3a that the undoped TiO nanocatalyst (TCNS0) showed
2
a strong absorption band around 380 nm in the UV region. But the TCNS sample
showed absorbance at 400–470 nm, with band expansion (about 100 nm) towards
the visible region. This expansion in the absorption edge decreases the direct band
gap of TCNS catalyst compared to undoped TiO (TCNS0) and this may be due to
2
Fig. 1 FTIR spectra of TCNS catalysts
123

906
B. Srinivas et al.
Fig. 2 SEM images of: a TCNS0 fresh, b TCNS5 fresh, and c used TCNS5
1
23

Selective solar photocatalytic oxidation of ethylbenzene
907
the insertion of C, N, and S into the TiO lattice [14, 21, 22]. Furthermore, the band
2
expansion in the DRS band increases with the increase in doped elements content
into the TiO lattice. The band gap energy (Eg value) of all of the catalysts is
2
estimated from the plot of absorbance versus photon energy. The absorbance is
extrapolated in order to obtain the band gap energy for the TCNS catalyst with good
approximation, as observed in Fig. 3b. The estimated band gap energies of TCNS0,
TCNS1, TCNS3, TCNS5, TCNS10, and TCNS15 are 3.05, 2.91, 2.82, 2.7, 2.6, and
2
.41 eV, respectively. From the DRS results, it is clear that the C, N, and S doping
.5
(
a) 4
TCNS15
TCNS10
TCNS5
TCNS3
TCNS1
TCNS0
4
.0
.5
.0
.5
.0
.5
.0
.5
.0
3
3
2
2
1
1
0
0
-0.5
400
500
600
700
800
Wavelength (nm)
(
b) 4.5
TCNS15
TCNS10
TCNS5
TCNS3
TCNS1
TCNS0
4
3
3
2
2
1
1
0
0
.0
.5
.0
.5
.0
.5
.0
.5
.0
-0.5
1.6
2.0
2.4
2.8
3.2
3.6
4.0
Bandgap (eV)
.5
(
c) 4
4
.0
.5
.0
.5
.0
.5
.0
.5
.0
3
3
2
2
1
1
0
0
Fresh
Used
-0.5
300
400
500
600
700
800
Wavelength (n.m)
Fig. 3 UV–Vis DRS of TCNS catalyst samples: a absorbance versus wavelength of thiourea doped
TiO , b absorbance versus band gap of thiourea doped TiO , c absorbance versus wavelength of TCNS5
fresh and used (third cycle)
2
2
123

908
B. Srinivas et al.
can cause band expansion for the absorption of TiO to the visible range and reduce
2
the band gap, which is beneficial for improving the photoabsorption and, ultimately,
the photocatalytic performance of TiO2.
XPS
To investigate the chemical states of the possible dopants incorporated into TiO2,
Ti2p, O1s, C1s, N1s, and S2p binding energies are studied by measuring the XPS
spectra [11].
Photocatalytic activity
Prior to the photocatalytic experiments, adsorption and photolysis studies were
carried out. The EB solution was kept in the dark without catalyst for one day and
-
1
-2
no oxidation is observed. 3.33 g L of the catalyst in 2.73 9 10 M of EB in
acetonitrile solution is kept in the dark. Aliquots were withdrawn at regular intervals
and the observation of no change in EB concentration indicates no adsorption of the
substrate on the catalysts prepared. The photolysis (without catalyst) experiment is
-
2
carried out under the solar light, taking 2.73 9 10 M of EB solution in the
Borosil glass tube and only 2% conversion was observed, with no desired product
even after 6 h of solar irradiation.
Selection of catalyst
To compare the TiO photocatalytic oxidation of EB activity of the commercial P25
2
and lab-made TiO (sol–gel method) samples and also CNS doped TiO (lab-made)
2
2
activity, evaluation runs were carried out under solar light irradiation.
-
1
All of the studies are carried out at 3 g L catalyst amount in 2.73 9 10
-2
M
EB in acetonitrile solution, keeping the molar ratio of EB to water as 5.4. The
photocatalytic activity of TCNS catalysts under solar light irradiation is shown in
Table 2. Among all of the samples, TCNS5 photocatalyst showed better activity.
The activity increased gradually with the increasing amount of dopent, and it
reached an optimum of 5 wt% loading of C, N, and S and any further increase
resulted in an activity decrease with time. The photocatalytic activity of the samples
can be attributed to the following factors. The work reported from our studies [11]
and others [23–25] regarding the doping of C, N, and S elements in titanium dioxide
results in visible light absorption spectral response in the photocatalytic samples and
the same is observed from the UV-DRS (Fig. 3). It can be seen from the DRS
spectra that C, N, and S doping resulted in an intense increase in absorption in the
visible light region and the absorption expansion of the base material titania
(
Fig. 3a, b). The band-gap narrowing of titania by C, N, and S doping lead to
enhanced photocatalytic activity of the titania under visible light. But we can also
see that, at higher loadings, the photocatalytic activity of TCNS samples decreased,
though they showed greater red shift in the absorption expansion. It might be due to
the fact that the excess dopent acts as recombination centers which facilitates
electron–hole recombination, thus, lowering the activity. So, the photocatalytic
1
23

Selective solar photocatalytic oxidation of ethylbenzene
909
activity is depressed to a certain extent. To conclude, the higher activity of the
TCNS5 sample can be ascribed to the high surface area, strong adsorption in the
visible region, and lower recombination of electron–hole pairs due to the high
concentration of surface hydroxyl groups, which can trap the photogenerated holes
and, thus, decrease the electron–hole recombination process.
Effect of illumination time on EB oxidation over TCNS5
In order to further study the process, the influence of reaction time on the catalytic
performance is illustrated in (Fig. 4). As the time progresses, the conversion of EB
is increased and reached a maximum of 56% by 12 h and, after that, no change in
concentration is noted due to the surface coverage of EB and, also, the products
formation reached a maximum. The concentration of acetophenone and 1-phenyl-
ethanol increases up to 6 h and, on further illumination, 1-phenylethanol decreases
with an increase of acetophenone under the same conditions. This behavior can be
explained by the dehydration of 1-phenylethanol after the subsequent attack of
hydroxyl radicals and is discussed in a later section of this paper.
Effect of EB concentration
In photocatalytic reactions, the observed rate constant decreases with the increase of
initial substrate concentration. The main steps during the photocatalytic reaction
occur on the surface of the catalyst, and, therefore, a high adsorption capacity is
associated, favoring the reaction. It is assumed that most of the reactions follow a
Langmuir–Hinshelwood kinetic rate model, which means that, for high initial con-
centrations, all catalytic sites are occupied. A further increase of the concentration
does not affect the actual catalyst surface-covered amounts, and, therefore, this
may result in a decrease of the observed rate constant. The effect of substrate
Table 2 Solar photocatalytic conversion of EB over: (1) P25 and TCNS (wt%), (2) TCNS0, (3) TCNS1,
(4) TCNS3, (5) TCNS5, (6) TCNS10, (7) TCNS15, (8) TCNS25, (9) TCNS40 doped TiO
2
catalysts
Catalyst
Conversion of
ethylbenzene (%)
Yield of
acetophenone (%)
Yield of
1-phenylethanol (%)
Others
(% yield)
a
TiO
2
P25
42.3
41
14
10
18.3
19
TCNS0
TCNS1
TCNS3
TCNS5
TCNS10
TCNS15
TCNS25
TCNS40
16
6
40.1
42.5
56.1
51.9
50.3
48.9
45.4
23.7
25.3
33.4
28.5
17.8
28.5
20
14.1
13.6
21.7
17.8
8.5
2.3
3.6
1.0
5.6
24
18.2
11.7
2.2
13.7
-
1
Reaction conditions: P = air 1 atm; T = 35 °C; irradiation time = 6 h; catalyst amount: 3.3 g L
illumination = solar light; EB: H O = 5.4
Benzaldehyde, benzoic acid, reductive dimerized products of benzaldehyde, acetophenone
;
2
a
123

910
B. Srinivas et al.
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
%
%
%
Conversion of Ethylbenzene
Yield of Acetophenone
Yield of 1-Phenylethanol
(a)
%
Yield others
5
0
0
2
4
6
8
10
12
Time (h)
Fig. 4 Conversion of EB time course during the oxidation over TCNS5. Reaction conditions: P = air
1
-1
atm; T = 35 °C; irradiation time = 6 h; catalyst amount: 3.3 g L ; illumination = solar light; EB:
a
H
2
O (molar) = 5.4. Benzaldehyde, benzoic acid, and diols
concentration is an important parameter for the photocatalytic conversion of EB
-
2
over known catalyst amounts. The concentrations of 0.34–16.38 (9 10 M) EB are
-
1
studied over TCNS5 catalyst with 3.33 g L catalyst amount. It is seen from the
observed data (Table 3) a decreasing trend in the conversion of EB when the initial
concentration is increased for the oxidation reaction. The concentration of
-
2
2
.73 9 10 M shows a maximum conversion and yield compared to other
concentrations. This indicates that it is optimum and, at higher concentrations, the
OH radicals produced by the catalyst are not sufficient to oxidize the EB molecules
-
2
which are adsorbed or near the catalyst surface. Hence, 2.73 9 10 M solution is
chosen for the oxidation, as there is an equilibrium between the adsorption of
reactant molecules and the generation of OH radicals from the active sites. The
-
2
concentration of EB below 2.73 9 10 M is converted into fragmented products
like benzoic acid, benzaldehyde, and benzene. When the concentration of EB is
-
greater than 2.73 9 10 M, 1-phenylethanol is converted into acetophenone
2
which, is further undergoing transformation as a reductive dimerized product like
2
,3-diphenyl-2,3-butanediol.
Effect of catalyst amount
In photocatalytic processes, the initial rates of reaction are found to be proportional
to the amount of catalyst. However, above a certain quantity, the reaction rate
becomes independent of the amount of catalyst. This limit corresponds to the
maximum amount of TiO at which all particles are totally illuminated. In order to
2
avoid an ineffective excess amount of the catalyst and to ensure a total absorption of
efficient photons on the surface, the optimum amount of catalyst has to be found.
1
23

Selective solar photocatalytic oxidation of ethylbenzene
911
Table 3 Effect of EB concentration over TCNS5 for photocatalytic oxidation
EB concentration
(
Ethylbenzene
(% conversion)
Acetophenone
(% yield)
1-phenylethanol
(% yield)
Others
(% yield)
-
9 10 M)
2
a
0
0
1
2
4
5
1
1
.34
.68
.36
.73
.09
.46
0.92
6.38
81.8
76.4
60
36.4
40.2
22.3
33.4
22.2
21.3
19.2
18.3
8.3
8.6
9.7
21.7
13.7
12.9
1
37.1
27.6
28
56.1
45.5
42.2
42
1
10
8
20
39.5
1.5
19.7
-1
Reaction conditions: P = air 1 atm; T = 35 °C; irradiation time = 6 h; catalyst: 3.3 g L ; illumina-
tion = solar light; EB: H O (molar) = 5.4
2
a
Benzaldehyde, benzoic acid, and diols
The data obtained in Table 4 illustrates the effect of variable activity of the
-
1
TCNS versus the catalyst quantities of 1.66, 3.33, 4.99, and 6.66 g L in the
suspension. It is observed that, increasing the catalyst amount from 1.665 to
-
.33 g L , the photocatalytic activity has increased, and at higher amounts, the
1
3
activity trend is not encouraging and the higher quantities of the catalyst makes the
solution turbid, which obstructs the light path into the solution and, in turn, reduces
-
1
the formation of OH radicals. In the present study, 3.33 g L is found to be the
optimum catalyst amount for the efficient oxidation of EB.
Effect of oxygen
The presence of an electron acceptor is essential in order to enhance the separation
of photo-generated electrons and holes, improving the efficiency of the photocat-
alytic process. We attempted to discover the influence of oxidation with air,
molecular oxygen, and nitrogen on EB oxidation. Air seems to be the more suitable
oxidant than molecular oxygen Table 5. The continuous flow of air or molecular
oxygen does not improve the extent of mild oxidation, but it leads to total oxidation,
Table 4 Effect of TCNS catalyst amount on the solar photocatalytic oxidation of EB
Catalyst
-
Ethylbenzene
(% conversion)
Acetophenone
(% yield)
1-phenylethanol
(% yield)
Others
(% yield)
1)
a
(
g L
1
3
4
6
.66
.33
.99
.66
54.25
56.1
53.6
52.9
8
15.5
21.7
0.2
12.7
1
33.4
26.3
26.9
27.1
25.8
0.2
Reaction conditions: P = air 1 atm; T = 35 °C; irradiation time = 6 h; illumination = solar light; EB:
O (molar) = 5.4
H
2
a
Diols, benzaldehyde, and benzoic acid
123

912
B. Srinivas et al.
like in any degradation/mineralization that occurs during photocatalytic treatment.
However, in the absence of an electron acceptor, the recombination of the electron–
hole pairs is very fast, inhibiting the photocatalytic process as observed from the
results obtained during N bubbling. The role of oxygen is discussed in the reaction
2
mechanism part. Dissolved oxygen is strongly electrophilic and, thus, an increase of
its content probably reduces unfavorable electron–hole recombination routes.
However, higher concentrations lead to a downturn of the reaction rate, which can
be attributed to the fact that the TiO2 surface becomes highly hydroxylated,
inhibiting the adsorption of substrate at the active sites [26]. Another point to make
from the XRD data [11], only anatase phase and no rutile phase is observed. The
increase in the oxidation of EB could be due to the oxygen-adsorbed capacity over
the TCNS catalyst, which contains only anatase. A further point to mention here is
that the rutile phase possesses a lower capacity to adsorb the oxygen than the
anatase phase [27]. In our study, the lesser activity of P25 is observed probably
because P25 is a combination of anatase and rutile.
Optimization of the EB:H O molar ratio
2
?
The investigation of water molecules for h scavenger in the presence of molecular
oxygen is illustrated. An interesting observation is that irradiation with solar light
provided high selective hydroxylation at the a-carbon to the benzene ring in EB.
The ratio of EB to water is evaluated in order to discover the role of water during
selective oxidation of the present substrate. The hydroxyl radicals generated from
water are thought to be attacked on a-carbon of the benzene ring than b-carbon in
EB. Molecular oxygen would be reduced by the photoexcited electron to form a
nucleophilic oxygen species such as superoxide. It is believed that some kind of
active oxygen would be generated in these photocatalytic oxidation systems, such as
superoxide, hydrogen peroxide, and hydroxyl radical, which would provide high
selectivity. For the direct and selective oxidation of the methylene group in EB,
selective generation of electrophilic active oxygen species in the system would be
important. This investigation resulted in the observation that undesirable oxidation
of side chains reduces the selectivity when we employ molecular oxygen and water
as an oxidant in sunlight irradiation. It can be seen clearly from Table 6 that, when
the molar ratio of EB to water increases, the conversion of EB is increasing and at
the particular ratio 5:4, the formation of acetophenone is found to be the highest.
Table 5 Influence of air/O
2
/N
2
on EB oxidation
Acetophenone
(% yield)
Oxidant
Ethylbenzene
% conversion)
1-phenylethanol
(% yield)
Others
(% yield)
a
(
N
2
2
2.0
58.0
56
0.5
32.0
33.4
–
1.5
6.5
1.0
O
19.5
21.7
Air
-
1
Reaction conditions: P = air/O
illumination = solar light; EB: H
2
/N
O (molar) = 5.4
Benzaldehyde, benzoic acids, and diols
2
1 atm; T = 35 °C; irradiation time = 6 h; catalyst: 3.3 g L ;
2
a
1
23

Selective solar photocatalytic oxidation of ethylbenzene
913
Further increase of water leads to total oxidation. It is speculated that EB is
adsorbed on the catalyst surface, radicals are generated at the methylene
carbon, with subsequent loss of a proton, and it attacks the hydroxyl forms in
1
-phenylethanol. The process continues until further attack of hydroxyl at the
methylene carbon. As two hydroxyl groups are existing on the same carbon, it loses
a water molecule in order to attain a stable state that leads to the formation of
acetophenone product.
Effect of particle dimensions and surface area
The lower photocatalytic activity of P25 (TiO ) compared with the lab-made sample is
2
clearly shown in Table 2. This can be attributed to particle size and the surface area of
P25 (TiO ) and TCNS0. Though the particle size of TCNS0 is less compared to P25,
2
the activity of EB oxidation is appreciable. This can be attributed to its solar harvesting
property compared to the P25 sample. Also from Table 1, the observed higher activity
of TCNS5 when compared with other CNS wt% loaded TiO catalysts are due to their
2
particle size and surface area. Particle size and, ultimately, particle dimensions are
responsible for the photoactivity of the semiconductor, since the electron–hole
recombination process was shown to be particle size-dependent [28]. P25 (TiO ) is
2
found to be less active for EB oxidation under solar light illumination. It is known that,
in the nanometer-size range, the physical and chemical properties of the semicon-
ductors are modified when compared to bulk size dimensions [29].
The particle size of 4.6 nm is found to be the optimum in the present case, and any
further reduction in size seems to decrease the activity, as is seen from Table 2. Small
variations in particle size leads to greater modifications in surface/bulk ratio
properties, thus, modifying the significance of volume and surface electron–hole
recombination processes. Generally, it is assumed that the lesser size the particle (in a
nanometer range), the higher the efficiency of the photocatalytic process. However, if
the TiO particle size is extremely reduced, in the range of a few nanometers, lattice
2
defects can lead to an increase in the electron–hole recombination.
Additionally, particle size is an important parameter for heterogeneous catalysis
in general, since it directly impacts the specific surface area of a catalyst.
The number of active surface sites increases with smaller particles and that results
Table 6 Effect of the EB:water molar ratio for photocatalytic oxidation over TCNS5 catalyst
Molar ratio
of EB:H
Ethylbenzene
(% conversion)
Acetophenone
(% yield)
1-phenylethanol
(% yield)
Others
(% yield)
a
2
O
1
2
4
5
6
.36
.71
.05
.40
.76
21.5
35.7
46.8
56.1
47.5
11.8
17.2
22.6
33.4
20.1
4.9
5.3
4.8
13.2
16.3
1.0
7.9
21.7
8.7
18.7
-1
Reaction conditions: P = air 1 atm; T = 35 °C; irradiation time = 6 h; catalyst amount: 3.3 g L
illumination = solar light
;
a
Benzaldehyde, benzoic acid, benzene, and diols
123

914
B. Srinivas et al.
in a higher rate of the photocatalytic reaction. Some studies revealed that the photo-
catalytic efficiency does not monotonically increase with decreasing particle size
and that there exists an optimal particle size of a few nanometers (3–10 nm) for pure
nanocrystalline TiO photocatalyst [30]. Though high thiourea loadings improved
2
the surface area, it has shown a detrimental effect in EB oxidation. These
observations are evident from Tables 1 and 2. The surface area of a solid catalyst is
directly related to the concentration of active sites for the adsorption and reaction. A
large surface area can be a determining factor in certain photocatalytic reactions,
since the adsorption of large mounts of substrate and oxygen promotes the reaction
rate. However, powders with high surface areas are usually associated with large
amounts of crystal lattice defects, facilitating the recombination of the photo-
generated electron–hole pairs, leading to a poor photocatalytic activity. It has been
reported that the photocatalytic activity of amorphous titanium dioxide is negligible,
indicating that crystallinity is an important requirement. Therefore, a balance
between surface area and crystallinity must be reached in order to obtain the
maximum photoactivity.
Effect of pH
The effect of pH is an important parameter because it commands the surface charge
properties of the catalyst and, therefore, the adsorption of the EB. The pH studies at
5
–10 are carried over TCNS5 catalyst. The adsorption capacity of the catalyst in
different pH ranges is not very much affected due to the non-ionic nature of EB. The
results depicted in Table 7 show that, at the original pH, the conversion of EB is
more or less the same compared to acidic or basic medium. This may be due to the
non-ionic nature of EB. In basic medium, there is a slight increase in 1-phenyl-
ethanol and this is observed when compared to the acidic medium. This may be
because the OH radicals groups and the hydroxylation at alpha carbon to the
benzene in EB, as well as in 1-phenylethanol, is clearly favored with an increase in
pH, whereas in acidic medium, such hydroxylation is not favored.
Catalyst recycling studies
To confirm the advantage of the present heterogeneous TCNS5 photocatalyst in the
liquid-phase oxidation process, we have reused the catalyst after separation from the
Table 7 Effect of pH for the oxidation of EB over TCNS5 photocatalyst
pH
Ethylbenzene
% conversion)
Acetophenone
(% yield)
1-phenylethanol
(% yield)
Others
(% yield)
a
(
5
55.5
55.5
56.1
23.9
24.9
33.4
15.5
19.5
21.7
16.1
11.1
1.0
1
7
0
-1
Reaction conditions: P = air 1 atm; T = 35 °C; irradiation time = 6 h; catalyst amount: 3.3 g L
illumination = solar light
;
a
Benzaldehyde, benzoic acid, and diols
1
23

Selective solar photocatalytic oxidation of ethylbenzene
915
Table 8 Recycle studies on photocatalytic oxidation of EB over TCNS5
Cycles
Ethylbenzene
% conversion)
Acetophenone
(% yield)
1-phenylethanol
(% yield)
Others
(% yield)
a
(
1
2
3
56.1
55
33.4
32
21.7
20
1
5
2
54
31
21
-1
Reaction conditions: P = air 1 atm; T = 35°C; irradiation time = 6 h; catalyst amount: 3.3 g L
illumination = solar light
;
a
Benzaldehyde, benzoic acid, and diols
reaction mixture followed by washing with acetonitrile while avoiding the loss of
catalyst particles. The results of reusability of the best catalyst are provided in
Table 8. After completion of the first cycle, the catalyst is recovered, dried, and
reused as such (without any calcination) for the second cycle. A slight decrease in
the conversion of EB is observed compared to the first cycle. When the same
catalyst is reused without calcination for the third cycle, a slight decrease in the
conversion of EB is observed compared to the first and second cycles.
The difference in activity is due to the accumulated organic intermediates on the
surface of the catalyst. These affect the adsorption concentration of the reactant
molecules, in turn reducing the activity. This observation is confirmed by calcining
the third cycle used sample at 400 °C for 3 h and subjecting it to a fourth cycle
activity evaluation. The original activity of the catalyst for EB conversion is
restored. This indicates that calcination of the used catalyst is necessary in order to
regain the activity. Furthermore, this is substantiated by comparison of the surface
characterization studies like SEM analysis (Fig. 2c of used catalyst), XRD (not
shown), and UV–Vis DRS techniques on the fresh and third cycle used samples
(
Fig. 3c). This confirms that the structural features of the TCNS5 did not change
during the oxidation. Thus, the above studies prove that the catalyst is reusable for a
number of cycles without any loss in activity and is stable for longer life. Based on
the results obtained, the following observations are made during EB oxidation with
air under soft reaction conditions. It is confirmed that the TCNS5 is stable and
reusable in the reaction. In view of this, the following mechanistic steps in the
oxidation of EB can be envisaged.
Plausible reaction mechanism
The photoexcitation of the semiconductor produces electrons generated in the
conduction band and holes in the valence band of TCNS catalyst on solar light
illumination:
hm
CNS ꢁ TiO2 ꢁ! CNS ꢁ TiO2ðe Þ + CNS ꢁ TiO2ðhþÞ
ꢁ
ð1Þ
ad
ad
The water molecule adsorbed on the surface of the catalyst reacts with the hole,
giving hydroxyl radical and a proton:
123

916
B. Srinivas et al.
þ
ꢂ
þ
H O ꢁ CNS ꢁ TiO + h ! OH + H
ð2Þ
Oxygen adsorbed on the catalyst surface serves as an electron acceptor for the
2
ðadÞ
2
conduction band electron that gives a super oxide radical anion:
CNS ꢁ TiO2ðeꢁÞ + O2 ! CNS ꢁ TiO2 ꢁ Oꢂ2ꢁ
ð3Þ
ad
ðadÞ
The proton generated in Eq. 2 reacts with the super oxide radical anion in Eq. 3,
giving rise to hydrogen peroxide (Eq. 4). This peroxide reacts with the photons,
producing two hydroxyl radicals:
ꢂꢁ
þ
ꢂ
2
O
+ H ! HO
2
hm
ð4Þ
ꢂ
ꢂ
2
HO ꢁ! H2O2 ꢁ! 2 OH
2
ꢁ
O2
Thus, the hydroxyl radicals formed react with the surface hydroxyl of the TiO2
surface which, in turn, loses a water molecule, resulting in a Lewis basic radical site
at the oxygen of the surface hydroxyl of TiO . The EB is reacted with Lewis acidic
2
center at the surface of the TiO . The subsequent loss of the proton at a-carbon to
2
the benzene in EB by Lewis basic free radical site facilitates to return the TiO into
2
its original form. Thus, the formed EB radical at a-carbon to the benzene in EB
reacts with the OH radical to give 1-phenylethanol. At this stage, any further
hydroxylation at low EB concentrations leads to the formation of benzaldehyde,
benzoic acid, benzene, and traces of reductive dimerized product of benzaldehyde.
The process repeats with hydroxyl radical and subsequent dehydration to form
Lewis basic radical on the catalyst surface. The proton at a-carbon to the benzene in
1
-phenylethanol reacts with the Lewis basic radical site, facilitating abstraction of
the proton from a-carbon to the benzene in 1-phenylethanol and, in turn, attacked by
hydroxyl radical to form dihydroxy a-carbon. At this stage, to attain the stability of
dihydroxy carbon, it loses a water molecule to form acetophenone releasing back
the TiO in its original form. The acetophenone formed can also undergo reductive
2
dimerization to form 2,3-diphenyl-2,3-butanediol, as depicted in Scheme 2. All of
the products are confirmed with mass and NMR spectroscopy and compared with
authentic samples.
Conclusions
The present study demonstrates C, N, and S doped TiO photocatalyst activity for
2
EB oxidation. The results conclude that 5 wt% thiourea doped TiO , i.e., TCNS5, is
2
an efficient catalyst and the high activity and selectivity may be due to the high
surface area, lower electron–hole recombination, and the stronger adsorption in the
-
visible light region. The substrate concentration of EB of 2.73 9 10 M, catalyst
2
-
1
amounts 3.33 g L , and a neutral pH are found to be favorable conditions for
higher oxidation rates. The catalyst activity is found to be sustainable even after the
third cycle (as evidenced by SEM, XRD, and UV–Vis DRS techniques). In
summary, the application of photocatalytic oxidation of EB offers a step forward in
sustainable catalysis. This is advantageous from the standpoint of low cost,
1
23

Selective solar photocatalytic oxidation of ethylbenzene
917
CH3
CH
CH3
OH
H
.
CH
O.
OH
.OH
Ti
O
Ti
Ti
O
Ti
Ti
O
Ti
-
H O
2
.OH
CH3
CH3
H
CH3
OH
C.
C
.
CH
OH
O
HO
O
HO
O
HO
.OH
Ti
Ti
Ti
Ti
Ti
O
Ti
-
H O
2
Further
hydroxylation
.OH
OH
COOH
CHO
CH3
OH
C
+
CH3
CH
HO
HO
Ti
O
Ti
-
H O
2
O
C
OH
CH3
CH3
CH3
Reductive
Ti
O
Ti
+
Dimerisation
OH OH
dl, meso
Scheme 2 A plausible reaction mechanism for EB oxidation over TCNS photocatalyst
environmentally friendliness, and operational simplicity, and it may be applicable to
large-scale reactions with further fine tuning of the system.
Acknowledgments The authors P.A.K.R. and M.S. thank the Council of Scientific and Industrial
Research (CSIR) (New Delhi, India) Emeritus Scientist Scheme.
References
1. P.P. Toribio, A. Gimeno-Gargallo, M.C. Capel-Sanchez, M.P. De Frutos, J.M. Campos-Martin,
J.L.G. Fierro, Appl. Catal. A 363, 32–39 (2009)
123

918
B. Srinivas et al.
2
3
4
5
6
7
. P.S. Singh, K. Kosuge, V. Ramaswamy, B.S. Rao, Appl. Catal. A 177, 149–159 (1999)
. C. Guo, Q. Peng, Q. Liu, G. Jiang, J. Mol. Catal. A 192, 295–302 (2003)
. H. Ma, J. Xu, Q. Zhang, H. Miao, W. Wu, Catal. Commun. 8, 27–30 (2007)
. S.K. Jana, Y. Kubota, T. Tatsumi, J. Catal. 247, 214–222 (2007)
. V.R. Choudhary, J.R. Indurkar, V.S. Narkhede, R. Jha, J. Catal. 227, 257–261 (2004)
. K.V. Subba Rao, B. Srinivas, M. Subrahmanyam, in Photo/Electrochemistry & Photo Biology in the
Environment, Energy and Fuel, ed. by S. Kaneco, B. Viawanathan, K. Funasaka (Research Signpost,
Trivendrum, 2004), p. 31–89
8
. G. Palmisano, V. Augugliaro, M. Pagliaro, L. Palmisano, Chem. Commun. 33, 3425–3427 (2007)
. A. Maldotti, A. Molinari, R. Amadelli, Chem. Rev. 102, 3811–3836 (2002)
9
1
0. O. Carp, C.L. Huisman, A. Reller, Prog. Solid State Chem. 32, 33–117 (2004)
1. P. Anil Kumar Reddy, P.V. Laxma Reddy, M. Sharma, B. Srinivas, V. Durga Kumari, M. Subrah-
manyam, J. Water Resour. Prot. 2, 235–244 (2010)
1
1
1
1
1
2. J. Madarasz, A. Braileanu, G. Pokol, J. Anal. Appl. Pyrolysis 82, 292–297 (2008)
3. X. Chen, S.S. Mao, Chem. Rev. 107, 2891–2959 (2007)
4. W. Ho, J.C. Yu, S. Lee, Chem. Commun. 10, 1115–1117 (2006)
5. M. Carraro, M. Gardan, G. Scorrano, E. Drioli, E. Fontananova, M. Bonchio, Chem. Commun. 43,
4
533–4535 (2006)
6. M.H. Habibi, A.Z. Isfahani, A. Mohammadkhani, M. Montazerozohori, Chem. Monthly 135,
121–1127 (2004)
1
1
1
1
1
2
7. J. Geng, D. Yang, J. Zhu, D. Chen, Z. Jiang, Mat. Res. Bull. 44, 146–150 (2008)
8. Y. Li, C. Xie, S. Peng, V. Lub, S. Li, J. Mol. Catal. A 282, 117–123 (2008)
9. S. Sakthivel, M. Janczarek, H. Kisch, J. Phy. Chem. B 108, 19384–19387 (2004)
0. Y. Yokosuka, K. Oki, H. Nishikiori, Y. Tatsumi, N. Tanaka, T. Fujii, Res. Chem. Int. 35, 43–53
(2009)
2
2
2
1. F. Dong, W. Zhao, Z. Wu, Nanotechnology 19, 365607–365611 (2008)
2. C. Gopinath, J.Phys. Chem. B 110, 7079–7080 (2006)
3. J.A.R. Herrera, K. Pierzchała, A. Sienkiewicz, L. Forro, J. Kiwi, J.E. Moser, C. Pulgarin, J. Phys.
Chem. C 114, 2717–2723 (2010)
24. Y. Amour, H.K. Shon, I.J. El Saliby, R. Naidu, J.B. Kim, J.H. Kim, Biosource Technol. 101,
1453–1458 (2010)
2
2
5. A. Zaleska, Physicochem. Probl. Miner. Process. 42, 211 (2008)
6. A.M. Braun, M.T. Maurette, E. Oliveros, in Photochemical Technology, ed. by J.W. Sons (Wiley,
Chichester, 1991), p. 397
2
2
2
3
7. Z. Zhang, C. Wang, R. Zakaria, J.Y. Ying, J. Phys. Chem. B 102, 10871–10878 (1998)
8. M. Herrmann, Top Catal. 34, 499–505 (2005)
9. X. Li, X. Quan, C. Kutal, Scripta Mater. 50, 499–505 (2004)
0. B. Ohtani, Y. Ogawa, S. Nishimoto, J. Phys. Chem. B 101, 3746–3752 (1997)
1
23