Sonication-assisted low-temperature routes for the synthesis of supported Fe-TiO2 econanomaterials: Partial photooxidation of glucose and phenol aqueous degradation

Article abstract of DOI:10.1002/cctc.201300025

Fe-doped TiO₂-supported photocatalytic materials prepared by the use of a mild ultrasound-assisted protocol have been found to possess excellent selectivities in glucose oxidation and total phenol mineralization. These materials were characterized by a number of techniques such as XRD, diffuse reflectance UV/Vis spectroscopy, N₂ physisorption, X-ray photoelectron spectroscopy, SEM with X-ray microanalysis, HRTEM, and flame atomic absorption spectroscopy. The proposed synthesis method was found to have a significant effect on textural properties, morphology, and visible-light responsiveness of nanophotocatalysts. The dopant agent was found to be in the form of Fe³⁺ ions. The effects of the textural properties, together with the Fe presence and the adsorption ability of the material (acidic properties), seem to be involved in the optimum photocatalytic activities found for Fe-TiO₂ supported on zeolite. In phenol photodegradation, although Fe doping demonstrated the detrimental effect on the conversion rate, an improvement in the direct mineralization of phenol to CO₂ and water, particularly for the SiO₂-supported photocatalyst, with a negligible content of toxic byproducts in water has been observed in comparison with commercially available Evonik P25.

Full text of DOI:10.1002/cctc.201300025

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201300025
Sonication-Assisted Low-Temperature Routes for the
Synthesis of Supported Fe–TiO2 Econanomaterials: Partial
Photooxidation of Glucose and Phenol Aqueous
Degradation
Juan C. Colmenares,*[a] Agnieszka Magdziarz,[a] Olga Chernyayeva,[a] Dmytro Lisovytskiy,[a]
Krzysztof Kurzydłowski,[b] and Justyna Grzonka[b]
Fe-doped TiO2-supported photocatalytic materials prepared by
the use of a mild ultrasound-assisted protocol have been
found to possess excellent selectivities in glucose oxidation
and total phenol mineralization. These materials were charac-
terized by a number of techniques such as XRD, diffuse reflec-
tance UV/Vis spectroscopy, N2 physisorption, X-ray photoelec-
tron spectroscopy, SEM with X-ray microanalysis, HRTEM, and
flame atomic absorption spectroscopy. The proposed synthesis
method was found to have a significant effect on textural
properties, morphology, and visible-light responsiveness of
nanophotocatalysts. The dopant agent was found to be in the
form of Fe3+ ions. The effects of the textural properties, to-
gether with the Fe presence and the adsorption ability of the
material (acidic properties), seem to be involved in the opti-
mum photocatalytic activities found for Fe–TiO2 supported on
zeolite. In phenol photodegradation, although Fe doping dem-
onstrated the detrimental effect on the conversion rate, an im-
provement in the direct mineralization of phenol to CO2 and
water, particularly for the SiO2-supported photocatalyst, with
a negligible content of toxic byproducts in water has been ob-
served in comparison with commercially available Evonik P25.
Introduction
The transformation of renewable biomass resources into valu-
able chemicals has gained significant attention in recent years.
Lignocellulose, the most abundant source of biomass, deriving
particularly from wood and agricultural residues, is nonedible
and can be used as a feedstock for the sustainable production
of chemicals.[1] Of various methods, which have already been
developed for biomass-reforming process (e.g., gasification,
fast pyrolysis, and supercritical conversion), high energy con-
sumption and low selectivity still remain significant prob-
lems.[2–4] From the viewpoint of sustainability, the development
of low-temperature and highly selective catalytic routes for
biomass transformation is required.[1]
cost, and availability) semiconductor materials are used.[6] Of
these, TiO2 is the most investigated semiconductor in photoca-
talysis because it combines unique and attractive characteris-
tics, such as high photocatalytic activity, stability, and environ-
mental tolerance.[7] Nevertheless, there are still important draw-
backs that severely limit a practical application of TiO2 photo-
catalysts, such as degrading organic pollutants in the gas or
liquid phase and/or performing useful transformations of or-
ganic compounds.[6] First, the band gap of the anatase phase
of TiO2, known as the most photoactive phase,[8] is too large
(3.2 eV) and does not enable an efficient absorption of most
sunlight.[9] In consequence, only approximately 5% of solar
light energy can be absorbed by TiO2. One of the efficient
methods to extend the absorbance of TiO2 in the visible region
is its doping with metallic elements. Of various transitional
metals, Fe is considered an appropriate candidate[10,11]: First,
because the size of the Fe3+ ion is similar to that of the Ti4+
ion, which makes the incorporation of Fe3+ into the crystal lat-
tice of TiO2 easier.[10] Second, because it favors the separation
of a photogenerated electron–hole pair.[8] Moreover, its impor-
tant implications in redox- and acid-catalyzed reactions,[12,13] as
well as in environmental remediation, have been reported be-
cause of its nontoxic nature, wide availability, and low price.[14]
Another important issue is to improve the recovery of the
powdered photocatalyst.[15] Therefore, it has been suggested
that fine TiO2 particles should be anchored on the supports for
ease of handling and for improving the adsorption and photo-
catalytic efficiency. Among various substrates, zeolites and SiO2
Herein, photocatalysis seems to meet these expectations as
it can be driven by sunlight and performed at room tempera-
ture.[5] For this process, special featured (photostability under
source of irradiation, chemical and biological inert nature, ca-
pability to adsorbed reactants under photonic activation, low
[a] Dr. J. C. Colmenares, A. Magdziarz, Dr. O. Chernyayeva, Dr. D. Lisovytskiy
Institute of Physical Chemistry
Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
Tel : (+48)223433215
E-mail: jcarloscolmenares@ichf.edu.pl
[b] Prof. K. Kurzydłowski, Dr. J. Grzonka
Faculty of Materials Science and Engineering
Warsaw University of Technology
Woloska 141, 02-507 Warsaw (Poland)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300025.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2270

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
stand out because of their pho-
tocatalytically attractive features,
such as good pollutant adsorp-
tion, diffusion properties, and
absence of light absorption
(transparency at 240 nm light
wavelength).[16]
Table 1. Structural and textural properties of the synthesized photocatalysts Fe–TiO2/ZeUI and Fe–TiO2/SiO2UI
in comparison with equivalent supports and TiO2/ZeUI and TiO2/SiO2UI.[a]
Material
XRD
N2 isotherms
Fe/Ti atomic ratio[b]
Nominal EDXRF FAAS XPS
[d]
[d]
[d]
Crystallite size[c] Crystal phase[c] SBET
Vp
dp
[nm]
[%]
[m2 gÀ1
]
[mLgÀ1
]
[nm]
Fe–TiO2/ZeUI
Fe–TiO2/SiO2UI 5.1
TiO2/ZeUI
TiO2/SiO2UI
Zeolite Y (Ze)
SiO2
12.2
A (100)
A (100)
A (100)
A (100)
–
658
209
700
206
784
197
0.45
0.50
0.46
0.75
0.52
0.39
0.8
14.6 0.08
0.8
14.6
0.8
0.08
0.11
0.08
–
–
–
0.07 0.19
0.09 0.20
Here, the synthesis of nano-
meter-sized spherical TiO2 parti-
cles with narrow size distribu-
tion through the ultrasound-as-
sisted impregnation of metal (Ti,
Fe) organic precursors into well-
defined inorganic supports is re-
ported. It is known that ultra-
sonic irradiation provides rather
unusual reaction conditions be-
13.1
5.7
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
14.0
–
[a] Fe/Ti atomic ratios estimated on the basis of different techniques; [b] EDXRF=energy dispersive X-ray fluo-
rescence; FAAS=flame atomic absorption spectroscopy; XPS=X-ray photoelectron spectroscopy; [c] A denotes
anatase; [d] Specific surface area (SBET), cumulative pore volume (Vp), and median pore width (dp) were calculat-
ed by the Barrett–Joyner–Halenda formula in the case of mesoporous SiO2-supported samples and the Hor-
vath–Kawazoe formula for microporous zeolite-supported samples.
cause of the cavitation phenomenon (formation, growth, and
implosion of bubbles inside the solution, which produce ex-
tremely high temperatures and pressures) that cannot be ach-
ieved by other methods. More uniform dispersion of nanopar-
ticles, higher surface area, better thermal stability, and phase
purity are some of the advantages of the sonication
method.[17] Therefore, ultrasound treatment has already been
coupled with different synthesis techniques,[18–20] such as a wet
impregnation method.[21,22] However, in these methods, ultra-
sonic irradiation has always been applied before solvent evap-
oration. Herein, in the proposed environmentally friendly
methodology, an improved ultrasonic treatment is also imple-
mented during the solvent evaporation step.
The anatase phase in the Fe–TiO2/ZeUI sample has also been
confirmed by TEM measurements and by the calculation of the
separation between lattice fringes (Figure 1a). The value was
estimated to be 3.48 ꢁ, and it corresponds to the lattice spac-
ing (101) of anatase TiO2. The same image indicates that the
oxide particles are attached to the external framework of the
zeolite substrate. However, some different TiO2 distributions
have also been observed. The zeolite surface covered well by
TiO2 particles can be seen in Figure 1b. A difference could also
be observed in the particle morphology; for example, the par-
ticles grafted on the zeolite external surface (Figure 1b) are
slightly elongated in comparison with the particles creating ag-
glomerates (results not shown here) that are spherical. The
TiO2 particle size has been estimated to be approximately
10 nm, which correlates well with the XRD measurements
(ꢀ12 nm). SEM with X-ray microanalysis (SEM-EDS) revealed
the presence of Ti, O, Fe, Si, and Al elements (Figure 1c). Al-
though the impregnation method involves a simple experi-
mental route, it usually provides broad particle size distribution
because it is difficult to control the size and shape of the parti-
cles by using this methodology. However, the improved wet
impregnation method coupled with ultrasonic irradiation pro-
posed here helped obtain a narrow-sized particle distribution
and mostly spherical particles.
The selected chemical transformations, as a proof of concept
for photocatalytic behavior, include the liquid phase selective
photooxidation of glucose to glucaric acid (GUA) and gluconic
acid (GA) as well as the total photocatalytic mineralization of
phenol in water under mild conditions.
Results and Discussion
Characterization
The XRD patterns (results not shown here) for the synthesized
samples demonstrate that only the anatase phase of TiO2 is
present in both materials. No reflections corresponding to Fe
were identified, probably because of the low Fe concentration
in our samples, which is below the detection limit of this tech-
nique, or because of good dispersion of the metal particles.
Another reason is often ascribed to the insertion of Fe3+ ions
into the TiO2 structure because of the similar ionic radius of
Fe3+ and Ti4+ (0.64 and 0.68 ꢁ, respectively),[11] which leads to
the formation of Fe–TiO2 solid solutions on the lattice sites of
TiO2.[8,15] The crystallite size of TiO2 nanoparticles calculated
from the Scherrer equation was found to be 12.2 nm for
Fe–TiO2/ZeUI and 5.1 nm for Fe–TiO2/SiO2UI (Table 1). The XRD
patterns also demonstrate that the TiO2 diffraction peak in the
SiO2 sample was broader than the corresponding peak in the
zeolite sample, which indicates its smaller size. It suggests that
the SiO2 matrix restrains the growth of TiO2 crystallites.
On TEM images for Fe–TiO2/SiO2UI, an amorphous phase of
SiO2 was visible (results not shown here). The presence of the
crystalline phase of TiO2 was confirmed by high-resolution
scanning TEM measurements (results not shown here). The
SEM-EDS spectrum and mapping of the individual elements
are shown in Figure 1. Mapping analysis (Figure 1d i and ii)
confirmed the presence of Si and Ti elements. The agglomera-
tion of SiO2 nanoparticles and quite good dispersion of TiO2
can be seen clearly. The Fe element was not visible by using
this technique, probably because of the low concentration;
however, SEM-EDS analysis (Figure 1e) and elemental analyses
(Table 1) confirmed its existence in the sample.
Bulk elemental analysis was performed by using energy dis-
persive X-ray fluorescence (EDXRF) and flame atomic absorp-
tion spectroscopy techniques. The Fe/Ti atomic ratios evalu-
ated from both techniques are indicated in Table 1, and they
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2271

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 1. TEM images of the Fe–TiO2/ZeUI photocatalyst: a) high-resolution scanning TEM image of TiO2 particles attached to the edge of the zeolite sub-
strate; b) SE image of the zeolite substrate covered with oxide particles; c) SEM-EDS spectrum. TEM images of the Fe–TiO2/SiO2UI photocatalyst: d) High-angle
annular dark-field-scanning TEM image of TiO2 particles on the SiO2 substrate, with i) mapping of Ti and ii) mapping of Si; e) SEM-EDS spectrum.
are congruent with nominal calculations, which imply the
quantitative incorporation of Fe into the TiO2-supported sam-
ples by using the proposed impregnation method coupled
with ultrasound methodology. In addition, the Fe/Ti atomic
ratio has been calculated by using X-ray photoelectron spec-
troscopy (XPS) on the surface of the materials, which indicates
that the surface is Fe-enriched (Fe/Tiꢀ0.2) compared to the
bulk TiO2.
the transition metal ions into the band gap of TiO2. Therefore,
the absorption edge shift toward longer wavelengths for the
Fe-doped TiO2 results from the electronic transition from the
dopant energy level to the conduction band of TiO2.[10]
The textural properties of the synthesized materials are sum-
marized in Table 1. Loading of TiO2 particles on zeolite and
doping with Fe ions resulted in the reduction of the surface
area and pore volume of the material. Thus, TiO2 nanoparticles
probably do not enter the pores of zeolite and they are more
likely embedded in the external surface area of the support
(result also confirmed by HRTEM measurements; Figure 1b). A
decrease in the specific surface area can be attributed to the
growth of TiO2 particles.[25] In the case of loading of TiO2 parti-
cles on SiO2 and doping with Fe ions, the surface area and
pore volume of the material increase slightly. Thus, the aggre-
gation of TiO2 particles on the SiO2 surface and the develop-
ment of voids in this aggregation could create an additional
pore volume.[26]
The diffuse reflectance UV/Vis spectra (Figure 2) of the syn-
thesized samples showed the extension of the absorption
band to the visible region and a significant enhancement of
light absorption at a wavelength of 400–500 nm as compared
to non-Fe-doped samples. The band-gap energies (Eg) of
Fe–TiO2/ZeUI and Fe–TiO2/SiO2UI calculated by Kubelka–Munk
theory[23] are 2.71 and 2.64 eV, respectively (Figure 2, inset). Ac-
cording to the previous report of Choi et al.,[24] the presence of
metal ions in TiO2 does not modify the position of the valence
band edge of TiO2. Instead, it introduces new energy levels of
XPS analysis was used to determine the surface composition
of the photocatalyst elements and their electronic states. The
survey spectra (Figure 3) confirmed the presence of Ti, O, Fe,
and Si in both samples. The Fe 2p3/2 and Ti 2p3/2 binding ener-
gies (BEs) corresponding to oxidation states are listed in
Table 2. (For comparative purposes, the XPS results of photoca-
talysts prepared without sonication treatment are presented in
Table S1.) The BEs of 458.54 eV for Fe–TiO2/ZeUI and 458.79 eV
for Fe–TiO2/SiO2UI correspond to the value of Ti4+ in the TiO2
lattice. In Fe–TiO2/SiO2UI, a small amount of Ti3+ is also ob-
served (Ti4+/Ti3+ =74:26, surface atomic percent ratio; Ti3+
was not observed for the catalysts prepared without sonication
treatment, as shown in Table S1). The BEs of 710.43 eV for
Fe–TiO2/ZeUI and 710.02 eV for Fe–TiO2/SiO2UI should be as-
signed to Fe3+, as they are close to 710.7 eV in a-Fe2O3.[27]
Figure 2. UV/Vis absorption spectra and band-gap values (Eg) of Fe–TiO2/
ZeUI and Fe–TiO2/SiO2UI compared with those of TiO2/ZeUI and TiO2/SiO2UI.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2272

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
The signal for O 1s showed four contributions (Table 2). The
first one, located at approximately 530 eV, corresponds to TiÀO
bonds; the second one, located at approximately 531–532 eV,
corresponds to FeÀO bonds. The third strong signal, at 533 eV,
relates to SiÀO bonds, and the fourth one, at 534.5 eV, corre-
sponds to CÀO bonds.
Photocatalysis
Liquid phase selective photooxidation of glucose
Two carboxylic acids were the main reaction products in the
liquid phase: GUA and GA. These carboxylic acids are impor-
tant building blocks for pharmaceutical, food, perfume, or fuel
industries.[5] Moreover, CO2 and traces of hydrocarbon species
(methane, ethane, and acetaldehyde) were detected. The selec-
tivities obtained for the studied photocatalysts Fe–TiO2/ZeUI,
Fe–TiO2/SiO2UI, TiO2/ZeUI, and TiO2/SiO2UI at the solvent com-
position 10% H2O/90% acetonitrile (ACN) are compared in
Figure 4. As shown in Figure 4, the incorporation of Fe helps
obtain more selective systems for the production of carboxylic
acids. The addition of ACN has been reported to improve se-
lectivity because of the “shield effect” driven by the ACN sol-
vent.[28] In the photocatalytic process of epoxidation of light al-
kenes, ACN (a weak base with pKa 25) stabilizes the reaction
products through solvation, which inhibits the proton transfer
while preventing the formation of radical species that afford
undesirable oxidation.[7] As shown in Figure 4, zeolite-support-
ed Fe-doped TiO2 is more selective to carboxylic
Figure 3. X-ray photoelectron spectroscopy survey spectra of a) Fe–TiO2/
ZeUI and b) Fe–TiO2/SiO2UI photocatalysts. Insets: XPS spectra of Fe 2p
region.
acids than SiO2-supported Fe-doped TiO2 at each
Table 2. X-ray photoelectron spectroscopy surface composition and electronic states
for Fe–TiO2/ZeUI and Fe–TiO2/SiO2UI photocatalysts.
time. The appreciable selectivity for the zeolite-sup-
ported photocatalyst is obtained in the first 5 min of
illumination (total selectivity for GUA+GA is 95.9%
with 7% glucose conversion), whereas for the SiO2-
supported photocatalyst, no selectivity in the liquid
phase is observed (the predominant product was
gaseous CO2) at the same time. Hence, better selec-
tivity properties are ascribed to zeolite-supported
photocatalysts. No significant difference exists in
glucose conversion between studied photocatalysts
during the first 20 min of the reaction. However,
after 180 min of the reaction, the difference in glu-
cose conversion is notable (100% for Fe–TiO2/SiO2UI
BE [eV] and Suggested species [%]
Photocatalyst
Fe–TiO2/ZeUI
Ti2p3/2
Fe2p3/2
O1s
458.54 Ti4+ (100) 710.43 Fe3+ (70)
529.73 TiÀO (15)
714.51 Fe3+ oxalate (30) 531.15 FeÀO (17)
532.78 SiÀO (57)
534.51 CÀO (11)
Fe–TiO2/SiO2UI 457.34 Ti3+ (26)
458.79 Ti4+ (74)
710.02 Fe3+ (63)
713.98 Fe3+ oxalate (37) 531.94 FeÀO (26)
533.12 SiÀO (46)
530.47 TiÀO (15)
534.79 CÀO (13)
Some other peaks at 714.51 eV for Fe–TiO2/ZeUI and 713.98 eV
for Fe–TiO2/SiO2UI correspond to Fe 2p3/2 and could be as-
signed tentatively to Fe3+[15]; the BEs of Fe3+ ions demon-
strates a positive shift (ꢀ714.0 eV) compared to those in
a-Fe2O3 (710.7 eV for 2p3/2). This probably indicates the forma-
tion of the oxalate-derived Fe salt (Fe2(C2O4)3·6H2O) with
a peak at 713.6 eV.[27] The appearance of that oxalate-derived
Fe salt is due to the interaction of Fe3+ with the surface
carbon (originated from organic precursors used during the
catalytic synthesis) detected by using XPS, particularly as oxa-
late anions (C2O42À), which are good chelating agents for metal
cations in the solution. Further investigations based on the ex-
tended X-ray absorption fine structure and Mossbauer spec-
troscopies will confirm these speculations.
and 87% for Fe–TiO2/ZeUI). Therefore, better degradation
properties are generally ascribed to the SiO2 materials, and
their photocatalytic behavior is due to the presence of Ti3+
ions on the surface (detected by XPS measurements; Table 2).
Ti3+ was not observed for the catalyst prepared without soni-
cation treatment (Table S1). In aqueous suspension systems,
electrons are trapped at surface defect sites (such as Ti3+) and
these defects can increase the degradation activity of the pho-
tocatalysts by retarding the electron–hole recombination.[29]
A
similar photocatalytic behavior is observed for the Fe-free pho-
tocatalysts TiO2/ZeUI and TiO2/SiO2UI (Figure 4). Herein, SiO2-
supported TiO2 samples also present better degradation prop-
erties, whereas zeolite-supported TiO2 samples demonstrate
better selectivities toward carboxylic acids. After 180 min of
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2273

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 3. Effect of the solvent composition on glucose conversion (X) and
total selectivity (SS) toward carboxylic acids (glucaric acid and gluconic
acid) in the liquid phase with different solvent compositions[a] (conver-
sions and selectivities given in [%]).
Photocatalyst
H2O/ACN 10:90
H2O/ACN 50:50
H2O
X [%]
SS [%]
X [%]
SS [%]
X [%] SS [%]
Fe–TiO2/ZeUI
19.2
61.5
7.2
94.3
8.3
0.0
20 min illumination
Fe–TiO2/ZeUI
60.4
41.1
19.6
57.5
13.0 0.0
90 min illumination
[a] Reaction conditions: 150 mL of the mother solution, 2.8 mmolLÀ1 of
glucose, 150 mg of the photocatalyst, 308C, 1 atm, 125 W UV lamp.
interval reaction times: 20 and 90 min of illumination—are
compared in Table 3. Notably, as the amount of water increases
in the solvent mixture, the glucose conversion rate decreases.
It has also been observed by Lꢂpez-Tenllado et al. in the pho-
tooxidation of crotyl alcohol to crotonaldehyde.[30] They claim-
ed that the competition of water and crotyl alcohol for the ad-
sorption sites of the catalyst could be the reason for this de-
crease. In the case of selectivity toward carboxylic acids, no se-
lectivity has been observed for 100% H2O during the whole re-
action. The best selectivity results have been obtained for 50%
H2O/50% ACN (94.2 and 57.7% after 20 and 90 min of illumi-
nation, respectively). Such a H2O/ACN volume ratio has also
been demonstrated by the mixture, which gives the highest
selectivity for bulk TiO2, as reported previously.[31] In the mix-
tures containing more than 10% H2O, the Fe-doped TiO2-sup-
ported zeolite photocatalyst is slightly more selective toward
gluconic acid (results not shown).
The photocatalytic properties of zeolite-supported materials
depend on the surface charge of zeolite, as the adsorption of
substrates/products on the photocatalyst is an important
factor that affects the photocatalytic activity. Y-type zeolites
are negatively charged materials; therefore, they adsorb on
the surface, by electrostatic attraction, cationic substrates and
repulse anionic substrates.[32] This behavior can facilitate the
selective photocatalytic oxidation of glucose because carboxyl-
ic acids under the reaction conditions (pH >2, pKa(GA) 3.35,
and pKa(GUA) 2.99) are repulsed but prevent their complete
photooxidation. The repulsion of carboxylic acids from the sur-
face is also observed for Fe–TiO2 supported on SiO2 (the same
reaction products); however, higher selectivity toward carbox-
ylic acids in the case of the Fe–TiO2/ZeUI photocatalyst indi-
cates the appropriate acidic character of the zeolite support
that helps in the selective photooxidation process. An et al. re-
ported the effect of the support acidity on better selectivity
toward gluconic acid in the oxidation of cellobiose.[33] Our pre-
liminary studies of photocatalysts by using temperature-pro-
grammed desorption of ammonia (Figure S2) have showed
two surface acidic centers (at 212 and 3618C) for Fe–TiO2/ZeUI
compared to only one type of acidic center (at 2118C) for
Fe–TiO2/SiO2UI. The synergetic effect of strong (3618C) and
weak (2128C) acidic sites of the zeolite-based photocatalyst
could help obtain more effective selectivity toward carboxylic
Figure 4. Selectivity and conversion results for the photooxidation of glu-
cose. Reaction conditions: 150 mL of the mother solution, 2.8 mmolLÀ1 of
glucose, 150 mg of the photocatalyst, 308C, 1 atm, 125 W UV lamp, H2O/
ACN 10:90 (v/v).
the reaction, only gas phase products (mostly CO2) have been
detected and the total mineralization of glucose have been
achieved by these systems.
Therefore, the solvent (H2O/ACN) optimization studies have
been conducted for Fe-doped TiO2-loaded zeolite. The selectiv-
ity and conversion results for different solvent mixtures—10%
H2O/90% ACN, 50% H2O/50% ACN, and 100% H2O after two
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2274

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
acids. In parallel to these preliminary studies, we have ob-
served a more negative zeta potential (ZP=À10 mV) for the
zeolite-based catalyst than for the SiO2-based catalyst (ZP=
À6 mV), which confirms the better repulsion of carboxylic
acids from the surface of the former catalyst (considerable
GA+GUA selectivity improvement). Moreover, the presence of
ACN in the solvent mixture helps stabilize carboxylic acids
through solvation and suppress their further oxidation.
tocatalysts, the amount of these compounds is negligible (Fig-
ure 5b).
An Fe leaching test has also been performed for the studied
reactions. No leaching was observed after both photocatalytic
tests. EDXRF measurements after 240 min of the photocatalytic
oxidation of glucose showed an Fe/Ti atomic ratio of 0.11 was
not observed in the first reaction and an Fe/Ti atomic ratio of
0.10 in the repeated reaction for Fe–TiO2/ZeUI. In the photoca-
talytic degradation of phenol, an Fe/Ti atomic ratio of 0.09 was
obtained for Fe–TiO2/SiO2UI. These values are comparable with
those of fresh photocatalysts (Table 1). In addition, the com-
plete absence of Fe and Ti ions in the liquid phase after the re-
action (240 min of illumination) was also confirmed by EDXRF
measurements.
Aqueous phase photodegradation of phenol
The photocatalytic activities of synthesized samples have also
been investigated in another test reaction: phenol degradation
in water. To verify the photocatalytic abilities of the samples
synthesized by our methodology, experiments were conducted
during 4 h of illumination. No appreciable phenol degradation
was found in the absence of UV irradiation or the catalyst. No
appreciable difference in the rate of phenol degradation (Fig-
ure 5a) is achieved for the studied photocatalysts, which have
reached approximately 20% after 240 min of illumination.
From the comparison of phenol degradation results under
the same reaction conditions obtained for the systems without
Conclusions
Fe-modified TiO2 photocatalysts were synthesized by using
a simple wet impregnation method coupled with ultrasonic ir-
radiation. The application of sonication, also during the solvent
removal step, produced spherical nanomaterials with good ele-
mental distribution. Materials were found to be photocatalyti-
cally active in the selective par-
tial oxidation of glucose to car-
boxylic acids (glucaric acid and
gluconic acid). The most selec-
tive system was the zeolite-sup-
ported photocatalyst (Fe–TiO2/
ZeUI), which after 20 min of illu-
mination achieved a selectivity
of 94.3% for a solvent composi-
tion of 50% H2O/50% ACN. It
has been found that reaction
conditions, particularly solvent
composition and short illumina-
tion times (5 min for Fe–TiO2/
Figure 5. a) Photocatalytic activity based on phenol decomposition (C/C0) as a function of time. Reaction condi-
tions: 150 mL of the aqueous solution, 150 mg of the photocatalyst, 50 ppm of glucose, 125 W UV lamp. b) High-
performance LC–MS representative analysis for 240 min of the photocatalytic degradation of phenol.
ZeUI), also have a considerable
effect on the activity/selectivity
of the tested photocatalysts. Fe
doping has been found detri-
Fe (ꢀ30% of degradation), it could be deduced that Fe
mental for the total oxidation of phenol in water. It has also
been suggested that the synthesis method prevails the forma-
tion of a-Fe2O3, which acts as a recombination center for pho-
togenerated charges and has a detrimental effect on the pro-
duction of OH radicals. It does not favor photodegradation
processes but improves the direct mineralization of phenol to
CO2 and water (particularly for the Fe–TiO2/SiO2UI photocata-
lyst) with a negligible content of toxic byproducts in water.
doping has a detrimental effect on the photocatalytic degrada-
tion rate in this reaction. This effect can be ascribed to the re-
ported fact that photoactivity can be reduced by introducing
Fe ions. When Fe is oxidized to a-Fe2O3, it remains at the inter-
face in contact with the TiO2 matrix and this oxide can act as
a recombination center of photogenerated charges.[34] The XPS
results indicate that the oxidation state of Fe in synthesized
photocatalysts is mainly Fe3+; thus, the synthesis method leads
to the formation of a-Fe2O3 (the most thermodynamically
stable Fe2O3 phase). Herein, the most active photocatalyst is
the commercially available Evonik P25, which achieved approx-
imately 70% of phenol degradation. Unfortunately, in the
aqueous phase, Evonik P25 leaves (Figure 5b) a high amount
of phenol-derived harmful compounds (hydroquinone,
1,4-benzoquinone, resorcinol, catechol, and 1,2,4-benzenetriol),
in addition to phenol, whereas in the case of synthesized pho-
Experimental Section
Titanium(IV) isopropoxide (TTIP, >98%, Acros Organics) was used
as a precursor of TiO2. Iron(III) acetylacetonate (99%, Acros Organ-
ics) was used as a metal precursor to prepare the Fe–TiO2
photocatalyst. Poly(ethylene) glycol (molecular weight 400, Acros
Organics) was used as a template direction agent. Y-type zeolite
(CBV 780, Zeolyst International) and SiO2 (AEROSIL 200, Evonik)
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2275

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
were used as supports. Glucose (ACS, pure p.a., POCH) and phenol
(>99%, Alfa Aesar) were used. Solvents and other reagents were
analytical grade and used without any further purification.
the dark for 20 min to reach a complete adsorption equilibrium of
glucose on the surface. A 125 W medium pressure mercury lamp
(lmax =365 nm) (Model RQ 3010, Photochemical Reactors Ltd.) was
used as a light source and placed inside the borosilicate glass im-
mersion well (300 nm, wavelength at which a 1 mm layer of the
glass absorbs at least 90% of the light). The reaction temperature
was 308C. Glucose solutions (2.8 mmolLÀ1) were prepared in
a Milli-Q water and ACN (10:90 v/v) mixture unless otherwise speci-
fied. Experiments were performed from the mother solution
(150 mL), and the catalyst with a concentration of 1 gLÀ1 was used.
All reactions were performed in ambient air (no oxygen bubbling
conditions). The samples (ꢀ2 mL) were taken from the photoreac-
tor at specified times and filtered through 0.2 mmꢃ25 mm nylon
filters to remove photocatalyst particles before HPLC analysis. Glu-
cose conversion and organic products were measured, after cali-
bration, with a Waters 590 HPLC
Synthesis of photocatalysts
Two photocatalysts modified with Fe were prepared: Fe–TiO2 sup-
ported on fumed SiO2 AEROSIL 200 (denoted as Fe–TiO2/SiO2UI)
and Fe–TiO2 supported on zeolite CBV 780 (denoted as Fe–TiO2/
ZeUI). Zeolite Y and SiO2 before the synthesis method were ther-
mally treated at 4508C during 8 h in static air. A simple wet im-
pregnation method coupled with ultrasonic irradiation was used to
synthesize these materials (Figure 6). The process began with the
addition of TTIP (5.35 mm) to the 2-propanol solution (70 mL) at
Pump equipped with a refractive
index detector (Waters 2414 Re-
fractive Index Detector). Separa-
tion was performed on an XBridge
TM
Amide Column (3.5 mm, 4.6ꢃ
150 mm) provided by Waters. The
mobile phase was Milli-Q water/
ACN (15:85 v/v) at a flow rate of
0.8 mLminÀ1
.
The
injection
volume was 10 mL. All reaction
products were identified by using
GC–MS (Hewlett Packard 6890
Series GC with 5973 MS, Hewlett
Packard-5 column) after silylation
Figure 6. Synthesis steps of the wet impregnation method coupled with ultrasonic irradiation.
by use of N,O-bis(trimethylsilyl)trifluoroacetamide with 1% tri-
methylchlorosilane as a derivatizating agent, performed at 608C
for 1 h in a CSB/COD-Reaktor ET 108. All products identified by
GC–MS were confirmed by using LC–MS (LC Prominence Shimadzu
coupled with MS 4000 Q-TRAP Applied Biosystems).
RT. Next, poly(ethylene) glycol (ꢀ2 g) and zeolite (2.5 g) were
added to the 2-propanol–TTIP mixture. The Fe precursor solution
(0.5 mm of iron(III) acetylacetonate in 10 mL of 2-propanol) was
added and the mixture was treated with ultrasonic irradiation (op-
erating frequency 35 kHz, 560 W, Sonorex Digitec-RC, Bandelin) for
60 min. Then, the solvent was evaporated slowly by a rotary evap-
orator coupled with ultrasonic treatment (the water bath was re-
placed with the ultrasonic bath; Figure 6). The solid material was
dried further for 4 h at 1108C and finally calcined at 5008C for 4 h
under air flow (flow rate 30 mLminÀ1). The Fe content was adjust-
ed to a nominal mass percentage of 1 wt%, and the nominal quan-
tity of TiO2 was fixed at 15%. The synthesis of SiO2-supported sam-
ples was conducted in a similar way, and the amount of SiO2 AERO-
SIL 200 taken to this synthesis was 3 g. Proportions of chemicals
were preserved. For comparative purposes, a conventional wet im-
pregnation method was used for the synthesis of Fe–TiO2/SiO2 and
Fe–TiO2/Ze without sonication treatment (see the Supporting Infor-
mation for results).
Aqueous phase photodegradation of phenol
All photocatalytic tests with phenol were performed in the setup
described above (for the selective photooxidation of glucose). The
lamp was placed in a borosilicate glass immersion well to avoid
the autodegradation of phenol (phenol disappearance was ob-
served under UV irradiation up to 300 nm; Figure S1). Phenol solu-
tions (50 ppm) were prepared in Milli-Q water. The reactions were
performed in ambient air. Phenol degradation was measured, after
external standard calibration, with a Waters 590 HPLC Pump
equipped with a photodiode array detector. Separation was per-
TM
formed on an XBridge C18 Column (5 mm, 4.6ꢃ150 mm) provid-
ed by Waters. The mobile phase was Milli-Q water/methanol (65:35
v/v) with 0.1% of CF3COOH at a flow rate of 1 mLminÀ1. The injec-
tion volume was 10 mL. The control experiments—photolysis or
without light—were performed to confirm that these photocatalyt-
ic reactions depended on the presence of both light and the pho-
tocatalyst. The whole spectrum of the medium pressure mercury
lamp used here is shown in Figure S1. Notably, photolysis reactions
led to a low conversion in both test reactions in the case of boro-
silicate glass immersion well (Figure S1). In phenol degradation,
conversion was approximately 11% after 120 min of illumination,
and in the case of glucose, it was approximately 6% after 120 min.
In the case of a quartz immersion well, approximately 30% of glu-
cose and 100% phenol degradation was achieved after the same il-
lumination time. Dark experiments (in the absence of a lamp) led
to no conversion of the starting material after 2 h of the reaction.
Photocatalytic experiments
All photocatalytic tests were performed in a Pyrex cylindrical
double-walled immersion well reactor with a total volume of
450 mL. Moreover, if different solids were tested as photocatalysts
for a certain organic reaction, a reference material (typically Evonik
P25) was included in the study for the sake of comparison.
Liquid phase selective photooxidation of glucose
The batch reactor was stirred magnetically to obtain a homoge-
neous suspension of the catalyst. The suspension was first kept in
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2276

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[15] Ch. Wang, H. Shi, Y. Li, Appl. Surf. Sci. 2011, 257, 6873–6877.
[16] M. V. Shankar, S. Anandan, N. Venkatachalam, B. Arabindoo, V. Muruge-
san, Chemosphere 2006, 63, 1014–1021.
Acknowledgements
This research was supported by a Marie Curie International Rein-
tegration Grant within the 7th European Community Framework
Programme. This scientific work was financed from the 2012–
2014 science financial resources granted for the international co-
financed project implementation (project no. 473/7.PR/2012, Min-
istry of Science and Higher Education of Poland).
[17] J. H. Bang, K. S. Suslick, Adv. Mater. 2010, 22, 1039–1059.
[18] S. Y. Kim, T. S. Chang, Ch. H. Shin, Catal. Lett. 2007, 118, 224–230.
[19] M. Zhou, J. Yu, B. Cheng, J. Hazard. Mater. B 2006, 137, 1838–1847.
[20] L. Zhou, W. Wang, L. Zhang, J. Mol. Catal. A-Chem. 2007, 268, 195–200.
[21] F. Tomul, Appl. Surf. Sci. 2011, 258, 1836–1848.
[22] Q. Wu, J. Ouyang, K. Xie, L. Sun, M. Wang, Ch. Lin, J. Hazard. Mater.
2012, 199–200, 410–417.
[23] S. Sakthivel, H. Kisch, Angew. Chem. 2003, 115, 5057–5060; Angew.
Chem. Int. Ed. 2003, 42, 4908–4911.
Keywords: heterogeneous catalysis · nanostructures · phenol
photomineralization · selective photocatalysis · ultrasonic
[24] W. Y. Choi, A. Termin, M. R. Hoffmann, J. Phys. Chem. 1994, 98, 13669–
13679.
[25] Ch. Wang, H. Shi, Y. Li, Appl. Surf. Sci. 2012, 258, 4328–4333.
[26] Ch. C. Wang, C. K. Lee, M. D. Lyu, L. Ch. Juang, Dyes Pigm. 2008, 76,
817–824.
[27] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. D. Bomben, Handbook of X-ray
Photoelectron Spectroscopy (Ed.: J. Chastain), Perkin–Elmer Corporation
Physical Electronics Division, USA, 1992.
[28] Y. Shiraishi, T. Hirai, J. Photochem. Photobiol. C 2008, 9, 157–170.
[29] L. Xiong, F. Yang, L. Yan, N. Yan, X. Yang, M. Qiu, Y. Yu, J. Phys. Chem.
Solids 2011, 72, 1104–1109.
[30] F. J. Lꢂpez-Tenllado, A. Marinas, F. J. Urbano, J. C. Colmenares, M. C. Hi-
dalgo, J. M. Marinas, J. M. Moreno, Appl. Catal. B 2012, 128, 150–158.
[31] J. C. Colmenares, A. Magdziarz, A. Bielejewska, Bioresour. Technol. 2011,
102, 11254–11257.
[32] G. Zhang, W. Choi, S. H. Kim, S. B. Hong, J. Hazard. Mater. 2011, 188,
198–205.
[33] D. An, A. Ye, W. Deng, Q. Zhang, Y. Wang, Chem. Eur. J. 2012, 18, 2938–
2947.
[34] B. Xin, Z. Ren, P. Wang, J. Liu, L. Jing, H. Fu, Appl. Surf. Sci. 2007, 253,
4390–4395.
[1] M. Stçcker, Angew. Chem. 2008, 120, 9340–9351; Angew. Chem. Int. Ed.
2008, 47, 9200–9211.
[2] S. G. Li, S. P. Xu, S. Q. Liu, C. Yang, Q. H. Lu, Fuel Process. Technol. 2004,
85, 1201–1211.
[3] M. Watanabe, H. Inomata, K. Arai, Biomass Bioenergy 2002, 22, 405–410.
[4] S. Rapagna, N. Jand, P. U. Foscolo, Int. J. Hydrogen Energy 1998, 23,
551–557.
[5] J. C. Colmenares, R. Luque, J. M. Campelo, F. Colmenares, Z. Karpinski,
A. A. Romero, Materials 2009, 2, 2228–2258.
[6] A. M. Balu, B. Baruwati, E. Serrano, J. Cot, J. Garcia-Martinez, R. S. Varma,
R. Luque, Green Chem. 2011, 13, 2750–2758.
[7] A. Maldotti, A. Molinari, Top. Curr. Chem. 2011, 303, 185–216.
[8] C. A. Castro-Lꢂpez, A. Centeno, S. A. Giraldo, Catal. Today 2010, 157,
119–124.
[9] X. Zhang, L. Lei, Appl. Surf. Sci. 2008, 254, 2406–2412.
[10] T. Tong, J. Zhang, B. Tian, F. Chen, D. He, J. Hazard. Mater. 2008, 155,
572–579.
[11] K. B. Qi, B. Fei, J. H. Xin, Thin Solid Films 2011, 519, 2438–2444.
[12] A. M. Balu, A. Pineda, J. M. Campelo, K. Yoshida, P. L. Gai, A. A. Romero,
Chem. Commun. 2010, 46, 7825–7827.
[13] A. Gervasini, C. Messi, P. Carniti, A. Ponti, N. Ravassio, F. Zaccheria, J.
Catal. 2009, 262, 224–234.
[14] K. M. Cross, Y. Lu, T. Zheng, J. Zhan, G. McPherson, V. John, Nanotechnol-
ogy Applications for Clean Water (Eds.: N. Savage, M. Diallo, J. Duncan,
A. Street, R. Sustich), William Andrew Inc., Norwich, USA, 2009, p. 347.
Received: January 13, 2013
Revised: March 11, 2013
Published online on June 3, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2270 – 2277 2277