Photodeposition of Ni nanoparticles on TiO2 and their application in the catalytic ozonation of 2,4-dichlorophenoxyacetic acid

Article abstract of DOI:10.1016/j.molcata.2011.11.001

Different photochemical approaches have been investigated to prepare supported nickel nanoparticles to be used in the catalytic ozonation of 2,4-dichlorophenoxyacetic acid (2,4-D). Direct photochemical (λ = 365 nm, without TiO₂) and photocatalytic deposition in presence or absence of sensitizers (TiO₂, acetone or benzophenone) were employed. The characterization of the catalysts was carried out by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM). The photocatalytic deposition without any sensitizer resulted to be the most suitable method to obtain in very short irradiation time, Ni nanoparticles randomly distributed over the support with a particle size ranging from 7 to 15 nm. An optimization of this technique was performed by varying the mass of TiO₂ and the light intensity. According to the results of 2,4-D degradation, the catalytic ozonation with Ni/TiO₂ catalyst presented a slightly higher conversion than TiO₂ or ozonation alone. These results were explained in terms of the two phases (NiO/Ni) present on the surface of TiO₂ which favor the ozone decomposition forming OH radicals useful for 2,4-D degradation. A reaction pathway including all the intermediates formed during the direct attack of ozone and the produced OH radicals was proposed.

Full text of DOI:10.1016/j.molcata.2011.11.001

Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Photodeposition of Ni nanoparticles on TiO2 and their application in the catalytic
ozonation of 2,4-dichlorophenoxyacetic acid
Julia L. Rodrígueza,b, Miguel A. Valenzuelaa,∗, Francisco Polac, Hugo Tiznadod, Tatiana Poznyakb
a Lab. Catálisis y Materiales, ESIQIE – Instituto Politécnico Nacional, Zacatenco, 07738 México, D.F. Mexico
b Lab. Ing. Química Ambiental, ESIQIE – Instituto Politécnico Nacional, Zacatenco, 07738 México, D.F. Mexico
c Centro de Investigación en Materiales Avanzados, Miguel de Cervantes 120, 31109 Chihuahua, Chih., Mexico
d Universidad Nacional Autónoma de México, Centro de Nanociencias y Nanotecnología, Km. 107 Carretera Tijuana a Ensenada, C. P. 22860, Ensenada, Baja California, Mexico
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 June 2011
Received in revised form 13 October 2011
Accepted 1 November 2011
Available online 10 November 2011
Different photochemical approaches have been investigated to prepare supported nickel nanoparticles
to be used in the catalytic ozonation of 2,4-dichlorophenoxyacetic acid (2,4-D). Direct photochemical
(ꢀ = 365 nm, without TiO2) and photocatalytic deposition in presence or absence of sensitizers (TiO2,
acetone or benzophenone) were employed. The characterization of the catalysts was carried out by X-
ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy
(TEM). The photocatalytic deposition without any sensitizer resulted to be the most suitable method
to obtain in very short irradiation time, Ni nanoparticles randomly distributed over the support with a
particle size ranging from 7 to 15 nm. An optimization of this technique was performed by varying the
mass of TiO2 and the light intensity. According to the results of 2,4-D degradation, the catalytic ozonation
with Ni/TiO2 catalyst presented a slightly higher conversion than TiO2 or ozonation alone. These results
were explained in terms of the two phases (NiO/Ni) present on the surface of TiO2 which favor the ozone
decomposition forming OH radicals useful for 2,4-D degradation. A reaction pathway including all the
intermediates formed during the direct attack of ozone and the produced OH radicals was proposed.
© 2011 Elsevier B.V. All rights reserved.
Keywords:
Ni nanoparticles
Photodeposition
Titanium dioxide
Catalytic ozonation
2,4-D
1. Introduction
about the reactivity with each kind of catalyst is not clear yet.
In contrast, supported metal catalysts have been scarcely studied
There is a growing need for understanding the mechanisms for
catalytic ozonation to introduce this technique in water treatment
at industrial scale [1–3]. Current research has been divided into two
routes, homogeneous and heterogeneous catalytic ozonation. Con-
cerning the last route, several catalysts have been tested, including
metal oxides, metals on supports, minerals and activated carbon,
among others [3]. The catalytic ozonation can proceed by direct
molecular ozone reactions and by an indirect pathway forming
OH radicals after ozone decomposition. However, the ozonation
process alone has its limitations, and the most important is that
the total mineralization of toxic organic compounds is difficult
to achieve. Therefore, the presence of a catalyst in the ozonation
process improves and controls the ozone decomposition and OH
radicals formation leading to mineralization [3].
specifically in the effect of the preparation method on the degra-
dation performance. In particular, if supported metal catalysts are
employed in the catalytic ozonation of contaminants is thought that
two reaction mechanisms can occur. By the first mechanism, organ-
ics are adsorbed on the surface of the catalyst and then attacked
by ozone or hydroxyl radicals; by the second mechanism, both
ozone and organics adsorb on the surface of the catalyst and a redox
process takes place forming organic radicals that are subsequently
oxidized by hydroxyl radicals or ozone [1,3].
As is well known, the catalyst preparation method plays an
important role in the size, morphology and particle size distribution
[12]. In fact, during the last two decades, several synthetic methods
such as chemical, thermal, photochemical, radiation, among oth-
ers, have been used in the preparation of metallic or metal oxide
nanoparticles [13–16]. According to our previous works, a pho-
todeposition method ensures the formation of small Pt particles
(2–4 nm) supported on C or TiO2 with a narrow particle size distri-
bution [17,18]. In a general way, the photodeposition process takes
place when the chromophore of a complex absorbs a photon result-
ing in a photoexcited state, then, the complex may decompose by a
photoredox reaction to produce a solid metallic phase cluster able
to deposit over the substrate surface [19–23]. The photodeposition
In recent years, several researchers have reported on the
degradation of organic pollutants by the heterogeneous catalytic
ozonation using TiO2 [4–6], NiO unsupported and supported [7–9]
and other metal oxides [10,11], however, a general conclusion
∗
Corresponding author. Tel.: +52 55 57296000x55112.
E-mail address: mavalenz@ipn.mx (M.A. Valenzuela).
1381-1169/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.11.001

30
J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
method is reproducible and the reaction can be stopped at any time
by switching the lamp off. However, byproducts coming from the
solvent oxidation/decomposition reactions can be generated after
irradiation [24].
2.3. Characterization techniques
The specific surface area of the supported catalyst Ni/TiO2
and the TiO2 support were determined by the BET method of
nitrogen adsorption using a MicromeritisAutochem II 2920. Metal
content in the prepared catalyst was obtained using an atomic
absorption spectrophotometer (Perkin Elmer). XRD patterns were
obtained in a XˇıPert-P analytical diffractometer with a Cu source
(ꢀK˛ = 0.15418 nm) and operated at 40 kV and 35 mA. The instru-
ment was coupled with a nickel filter and a X’Celerator detector and
it was settled in tetha/2tetha configuration. The scanning angle (2ꢁ)
range was varied to 20–90◦ with 0.016 step size and 20 s of counting
time. Phase analysis was carried out matching diffracted intensities
with PDF cards.
TEM images were obtained using a JEOL-JEM-2200 field emis-
sion operated at 200 kV. The samples were prepared with the
catalyst (<1 mg) in methanol and dispersed by ultrasound for
5 min. Thereafter, a drop of the solution was placed over a car-
bon coated Cu grid (300 mesh) and dried at room temperature.
Photoelectron core-level spectra of the as-prepared samples were
obtained with an X-ray photoelectron spectroscopy (XPS) system
(Riber LDM-32), which has an analysis chamber equipped with
an ion pump, an electron-energy analyzer (Mac-3) and a dual
anode X-ray source. XPS data were collected using the Al K␣ line
at 1486.6 eV. The high resolution XPS scans were completed at
0.1 eV energy steps and 0.5 eV resolution (∼25 eV pass energy),
while the surveys at 1 eV steps and 3 eV resolution (∼100 eV
pass energy). The spectra were decomposed into their compo-
nents with mixed Gaussian–Lorentzian lines by a non-linear least
squares curve-fitting procedure, using the public software pack-
age XPSPEAK 4.1. The binding energies and FWHM of the peaks
were determined from the fitting results after subtraction of the
Shirley-type background. Deconvoluted peak areas and standard
sensitivity factors were used to evaluate the surface composition of
the samples.
Diverse catalysts based on supported nickel have been used in
chemical processes industrially important, including hydrogena-
tion, methanation, reforming, dechlorination, desulfurization, and
so on. These catalysts are generally synthetized by impregnation
or coprecipitation with nickel salts on a support [25]. Neverthe-
less, it presents the problem of formation of mixed nickel oxides,
especially for nickel loadings of 10 wt% and high calcination tem-
peratures (≥600 ◦C), which are required to completely reduce
the Ni(II) cations to metallic nickel particles, which leads to the
unwanted metal sintering [26]. Therefore, one of the main advan-
tages of the photodeposition method is the direct reduction of
the Ni precursor avoiding reductive thermal treatments. Works
devoted to the photodeposition of Ni nanoparticles on semicon-
ductors are scarce and the effect of synthesis parameters has not
been reported yet [27].
In this study, we have explored the use of TiO2 to promote
the photocatalytic deposition of Ni nanoparticles. The nickel (II)
present in the precursor solution is reduced with the photo-
generated electrons on the TiO2 surface. Additionally, an easily
oxidable compound (e.g. ethanol) must be present in the system
to assure that holes are consumed to avoid their accumulation
on the photocatalyst surface, which could stop the photocat-
alytic activity or cause nickel reoxidation [28,29]. It seems that
the addition of oxalate ion as a co-reactant favors the reduction
of nickel (II) on TiO2 by an indirect photoreduction in aqueous
solution [27].
The evaluation of our photodeposited Ni/TiO2 catalysts was car-
ried out in the degradation of herbicide 2,4-dichlorophenoxyacetic
acid (2,4-D) by heterogeneous catalytic ozonation in aqueous solu-
tion. 2,4-D was selected as a model compound, since it is the most
widely used in agriculture sector and herbicide in the world, it is
poorly biodegradable and has been detected as a major pollutant
in ground and surface water [30].
2.4. Ozone generation and ozone consumption control
2. Experimental
Ozone was generated from dry oxygen by the ozone genera-
tor (corona discharge type) HTU500G (AZCO Industries Limited
– Canada). The Ozone Analyzer BMT 964 BT (BMT Messtechnik,
Berlin) provides the ozone monitoring in the gas phase at the
reactor outlet for the control of the ozonation degree, the ozone
consumption and the ozone decomposition as well.
2.1. Materials and reagents
The bis(2,4-pentandionato) Ni(II) (Ni(acac)2, Aldrich) was
reagent grade. Anhydrous ethanol and acetone (J.T. Baker) were
spectrophotometric grade. TiO2 powder P25 (Degussa AG, 80%
anatase and 20% rutile, non-porous, BET surface area = 44 m2 g−1
)
was used as starting material. 2,4-dichlorophenoxyacetic acid (Alfa
Aesar, 90%) and benzophenone (Aldrich, analytical grade) were
used as received.
2.5. Ozonation procedure
All experiments with ozone were carried out in a semi-batch
type reactor (0.5 L) at 21 ◦C. The agitation was provided by means
of an ozone–oxygen mixture bubbling through a ceramic porous
filter, which is placed at the bottom of the reactor. The initial ozone
concentration was 25 mg L−1. The ozone–oxygen mixture flow was
0.5 L min−1. A flow diagram showing the ozonation procedure and
equipment is depicted in Fig. 1.
2.2. Catalyst preparation
In a glass reactor, it was added a TiO2 (0.1 g) suspension con-
taining a 8 × 10−4 M alcoholic solution of Ni(acac)2 at 25 ◦C to have
a total volume of 100 mL, purged with nitrogen. The mixture was
irradiated with 14 black light UVA lamps (8 W) with a maximum
emission at about 365 nm. The sensitized photoreaction was per-
formed in presence of acetone (0.2 M) or benzophenone (10−3 M).
The suspension was subjected to vigorous and continuous stirring
to avoid both the film formation and sedimentation of TiO2. After
irradiation, the sample was dried at 120 ◦C to evaporate the solvent.
The kinetics of the Ni(acac)2 photodecomposition was performed
by using a Lambda UV–Vis spectrophotometer (Perkin Elmer) at
wavelength of 310 nm.
The model solution of 2,4-D herbicide was prepared with a
concentration of 80 mg L−1. Aliquot of 3-mL ozonation reaction
solution was withdrawn at time intervals from the reactor for
sequent analysis. The catalyst concentration was 0.1 g L−1.A HPLC
apparatus (Perkin-Elmer series 200, UV/Vis detector) was used
to record the change of concentration of 2,4-D, under the fol-
lowing operation conditions: column Prevail Organic Acid 5 ␮
“Grace”, 150 × 4.6 mm with 60:40 acetonitrile–buffer mobile phase
of KH2PO4 at 25 mM adjusted a pH 2.6 with H3PO4 at wavelength
225 nm with the flow 1 mL min−1
.

J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
31
Fig. 1. Schematic diagram of the ozonation apparatus.
3. Results and discussion
photodecomposition of Ni(acac)2.
A
beneficial behavior was
observed with acetone and the decomposition rate is increased
in 18% in 6 h of irradiation time, while by using benzophenone
the same decomposition degree was obtained but just in 1 h. This
photosensitized reduction uses the photoactive reagents (e.g. ben-
zophenone,) generating intermediates after UV irradiation, which
reduce the metal precursor to its metallic form [24]. Furthermore,
when the photodecomposition reaction proceeds in presence of any
sensitizer, the solution color turned form light green to dark brown.
This change in the Ni(acac)2 solution color has been attributed to
the formation of nickel colloidal particles [21,36].
The kinetic analysis of the photochemical reaction with or with-
out sensitizers is also presented in Fig. 2. It is clearly shown that the
raw data fitted well to a first order kinetics. Note that in presence of
BP, the rate constant (k = 0.17 h−1) was one magnitude order higher
with respect to the rate constant in presence of AC (k = 0.03 h−1).
The photochemical reaction, without any sensitizer, was very low
with a rate constant of k = 0.0078 h−1. The improved reaction rate
obtained by the addition of BP may be due to the electron transfer
reaction between the sensitizer and the Ni precursor as shown in
the following reactions (Scheme 2) [37].
3.1. Photochemical reaction
The evolution of the Ni(acac)2 normalized concentration as a
function of irradiation time is compared in Fig. 2. As can be seen, a
negligible decomposition of Ni(acac)2 is observed after 6 h of irra-
diation time without the presence of a sensitizer, which means that
the photoexcitation of the Ni precursor did not occur at the wave-
length used in the experiment (ꢀ = 365 nm). However, in presence
of acetone or benzophenone, a mild conversion of the Ni precursor
is detected. Indeed, several researchers have reported that acetone
and mainly aromatic ketones improve the photochemical decom-
position of metal complexes by sensitization [31,32].
The photochemical reduction has already been reported and this
reaction can proceed by the following mechanism (Scheme 1) [33]:
The photodecomposition rate of Ni(acac)2 in ethanol at 365 nm
is very low because only 1% Ni(acac)2 is decomposed in 6 h (Fig. 2).
This low photodecomposition rate can be increased by using the
“acetone method”, which has been proved successfully in the
formation of Ag nanoparticles [34,35]. Therefore, two photosen-
sitizers acetone (AC) and benzophenone (BP) were tested in the
3.2. Photocatalytic reaction
1.0
0.8
0.6
The photocatalytic deposition of Ni nanoparticles, employing
Ni(acac)2 in ethanol solution (as sacrificial electron donor) was car-
ried out in presence of sensibilizer and without it, the results are
shown in Fig. 3. The presence of TiO2 in the reaction lead to a rapid
reduction of the Ni precursor compared with the photochemical
reduction. As it is known, the conduction band electrons on the sur-
face of a TiO2 particle has a reduction potential of −0.5 V (vs. NHE
at pH = 7), then, it can reduce Ni ions because their redox poten-
tial Ni2+/Ni is −0.24 V [38]. However, it has been suggested that an
overpotential of 0.3 V is required for an efficient reduction. As can
be seen in Fig. 3, the nickel reduction was at about 90% after 1.5 h
of irradiation time, and then no more reduction was observed. The
presence of either sensitizer presented a much lower conversion of
the Ni precursor, mainly at low irradiation times (0–2 h). Probably,
after the adsorption of BP and AC on the surface of TiO2 hinder the
photons absorption by the semiconductor.
Ni(acac)2
Ni(acac)2 - AC
Ni(acac)2 - BP
0.4
0.2
The dimensionless residual concentration of Ni(acac)2 as a func-
tion of the incident light intensity after 1.5 h of irradiation is shown
in Fig. 4. Note that the tendency of the concentration profiles with
or without sensitizer was similar, practically linear in all the range
of light intensity. Nevertheless, the photocatalytic deposition of Ni
nanoparticles on TiO2 without sensitizer was always higher. An
explanation of this behavior can be due to a competition for the
Time,h
0.0
0
1
2
3
4
5
6
7
Time, h
Fig. 2. Evolution of the dimensionless concentration of Ni(acac)2 with irradiation
time in presence of sensitizer (AC – acetone; BP – benzophenone) and with-
out it. Experimental conditions: [Ni(acac)2] = 8 × 10−4 M, [AC] = 0.2 M, [BF] = 10−3 M,
ꢀ = 365 nm. Insert kinetic analysis assuming a first order reaction.

32
J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
Scheme 1.
Scheme 2.
incident photons by the TiO2 and the sensitizer. Indeed, as the
wavelength of the lamps used in our experiments was centered
at 365 nm, then, higher photons absorption must take place with
the TiO2 than with the organics.
causing the reduction of Ni(II) adsorbed on TiO2. The tendency to
achieve a plateau in the kinetic curve in Section III could be indica-
tive that the reduction of Ni(II) was completed. As a consequence,
the intensity of light that penetrates deep into the TiO2 gradually
falls and the nickel photodecomposition did not present variations
with irradiation time.
The effect of the TiO2 concentration on the photoreduction of
Ni(acac)2 in ethanol was also studied and reported in Fig. 6. These
results indicated the expected trend: the higher the TiO2 concen-
tration, the higher conversion of the Ni(acac)2. According to these
results, an optimal concentration of TiO2 was reached at about
1 g L−1. Most studies reported that a enhanced degradation rates
are obtained for catalyst loads up to 0.4–0.5 g L−1 [32]. This behav-
ior can be explained in terms of the availability of active sites on the
TiO2 surface and the light penetration into the suspension. Indeed,
To clarify the behavior shown in Fig. 4, data concerning to the
evolution of the Ni (acac)2 as a function of irradiation time at
several light intensities are shown in Fig. 5. The curves exhib-
ited a reverse S-shaped form with three well-defined sections;
Section I corresponds to an induction period with the nickel pre-
cursor decomposition giving rise to a small absorption change. The
length of the induction period depends inversely on the light inten-
sity. During this time begins the photoreduction of Ni(II) complex
to intermediate species, including Ni(I) and Ni(0). The significant
shift in the Ni(acac)2 photodecomposition is observed in Section II
1.0
1.0
0.8
0.6
Ni(acac)2 - TiO2
Ni(acac)2 - BP - TiO2
Ni(acac)2 - AC - TiO2
0.8
0.6
0.4
0.2
0.0
0.4
Ni(acac)2 - TiO2
Ni(acac)2 - AC - TiO2
Ni(acac)2 - BP - TiO2
0.2
0.0
0
5
10
15
20
25
30
35
40
45
50
55
60
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Time, h
Intensity, mW cm-2
Fig. 3. Evolution of the dimensionless concentration of nickel reduction with
irradiation time at 365 nm. Experimental conditions: [Ni(acac)2] = 8 × 10−4 M,
Fig. 4. Evolution of the dimensionless concentration of the photocatalytic reduc-
tion of Ni(acac)2 as
a function of light intensity. Experimental conditions:
[AC] = 0.2 M, [BF] = 10−3 M, ꢀ = 365 nm, [TiO2] = 1.075 g L−1
.
[Ni(acac)2] = 8 × 10−4 M, ꢀ = 365 nm, [TiO2] = 1.075 g L−1
.

J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
33
9.00x10
7.50x10
6.00x10
4.50x10
3.00x10
58.8 W cm-2
46.3 W cm-2
29.4 W cm-2
I
16.8 W cm-2
II
III
1.50x10
0.00
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
Time, h
Fig. 5. Evolution of the Ni(acac)2 concentration as a function irradiation time with
respect a different light intensity. Experimental conditions: [Ni(acac)2] = 8 × 10−4 M,
ꢀ = 365 nm, [TiO2] = 1.075 g L−1
.
Fig. 7. Diffraction pattern of Ni/TiO2 sample.
there is a limit of catalyst loading in which agglomeration and sed-
imentation dominate and the catalyst surface became unavailable
for photon absorption and adsorption of the reactant. The optimum
loading of photocatalyst depends on the geometry of the reactor,
the type of UV-lamp, the operational conditions of the reactor and
of course on the initial solute concentration [39]. Note that the esti-
mated initial reaction rates against TiO2 agreed very well to a first
order kinetics in the inserted plot in Fig. 6.
value of Ni/TiO2 was 6.09 and for TiO2 was 6.46 so that the pres-
ence of nickel did not change significantly the isoelectric point. H2
chemisorption data show a dispersion of 6.5% which correspond
to an average particle diameter of 16.5 nm. Comparable values of
average Ni size diameters have been obtained from TEM analysis
(dTEM ≈ 13 nm).
Fig. 7 shows the XRD pattern of Ni/TiO2 synthesized by
the photocatalytic deposition method. The XRD pattern of
Ni/TiO2 sample evidence TiO2 rutile and anatase phases with
tetragonal structures, identified with PDF cards (73–1232) and
(84–1285), respectively. Lattice parameters of semiconductor
phases were a = b = 0.45936 nm and c = 0.29587 nm for rutile and
a = b = 0.37848 nm and c = 0.9512 nm for anatase. In the Ni/TiO2
sample, Ni0 phase was observed at 2ꢁ = 44.34◦, identified as (1 1 1)
plane of Ni cubic structure (a = 0.35238 nm) according to PDF card
(04-0850).
3.3. Catalyst characterization
The real nickel content on TiO2 was 3.6 wt%, compared with
a nominal amount of 5 wt%, which confirms that some of the
Ni precursor is maintained in solution without further reduction.
Nitrogenphysisorptionresults(notshownhere)indicatedthatTiO2
had major specific surface area (SSA 44 m2 g−1) compared with that
of Ni/TiO2 (33 m2 g−1), this reduction in SSA is due to the block-
age of the pores support during the deposition reaction. The IEP
Fig. 8 reports TEM images with different magnification of the
sample prepared by photocatalytic deposition. TEM micrographs
(Fig. 8a and b) show Ni aggregates deposited over the crystalline
TiO2 semiconductor. Aggregates are formed with Ni nanoparticles
ranging from 7 to 15 nm in size although not well defined geo-
metrical shape (see Fig. 8a). Apparently, small Ni nanoparticles are
formed in situ during the reaction and are well dispersed in alco-
holic media; however, during evaporation and sedimentation steps,
the particles aggregate fall down locally over TiO2. Then, the depo-
sition of Ni particles is not homogeneous over the semiconductor
surface. Micrograph of Fig. 8c shows regions of high concentrations
of Ni rather than individual metallic particles.
1.0
0.8
0.6
2.0x10
The chemical nature of the Ni-containing catalyst has been
investigated with XPS analysis. The spectra were charge corrected
using the Ti 2p3/2 peak position (459.5 eV for TiO2) [40], as the C 1s
signal was composed by several peaks. In a survey spectrum (not
shown) there were found signals only for Ni, O, Ti and C, as expected.
The corresponding high resolution regions are presented in Fig. 9.
The Ti 2p3/2 region can be fitted with one peak (FWHM 1.6 eV),
indicating a single Ti chemical specie. The C 1s peak is fitted with
three peaks. The first one at 284.5 eV is assigned to adventitious
carbon. The other two peaks at 286.0 and 288.4 eV are assigned
to organic carbon with increasing grade of oxidation [40–42]. The
presence of the organic carbon indicates that the reaction byprod-
ucts were not removed completely after drying of the sample
and/or incomplete Ni precursor decomposition. The 38% of the O 1s
0.4
0.0
0.0
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
[TiO2], g L-1
Fig. 6. Evolution of the dimensionless concentration of nickel reduction with
TiO2 loading and its relation with initial rate reaction. Experimental conditions:
[Ni(acac)2] = 8 × 10−4 M, ꢀ = 365 nm. Insert kinetic analysis.

34
J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
contribution of more O peaks, perhaps the extra O is bonded to the
organic carbon and to the Ni oxide found in the sample, see below.
However, with the current data resolution it is not possible to fur-
ther deconvolute this peak to propose the positions of the extra O
peaks.
The signal-to-noise ratio of the Ni 2p spectral region was poor
at the acquired resolution, despite the long period of data collec-
tion and does not allows for a reliable interpretation. In order to
improve the peak intensity, we opted to acquire a slightly lower res-
olution spectrum only for the Ni 2p region, which is shown instead
of the high resolution one. Nevertheless, it can be extracted valuable
information. There are two main peaks at 855.8 and 873.5 eV corre-
sponding to Ni 2p3/2 and 2p1/2, respectively. On the HBE side of each
main peak appears a satellite shoulder (approximately at 861.5 and
881.0 eV), which intensity, respect to the corresponding main peak,
is the distinctive feature of oxidized nickel, i.e. NiO [36,44,45]. Addi-
tionally, it can be seen a small shoulder approximately at 852.0 eV
corresponding to metallic Ni. Since the nickel oxide 2p region shape
is complex (due to the presence of two satellites, which shape is non
symmetrical), and given the lack of spectral resolution we consider
adventurous trying to deconvolute the peaks, and no atomic per-
cents for NiO or metallic Ni are calculated. Still, given the observed
intensities, NiO is the main component in the sample, and the over-
all nickel concentration is ∼6 atomic % in respect to titanium. Nickel
nanoparticles on TiO2 formed after photocatalytic reduction may
be accompanied by an oxidation process with atmospheric oxygen.
Indeed, in our catalysts it was evident this behavior and we detected
metallic nickel by XRD (Fig. 7) and NiO by XPS (Fig. 9d). These results
agreed with those reported for several supported nickel catalysts
where the uppermost surface of nickel nanoparticle is reoxidized
easily as shown by XRD and XPS studies [46,47].
3.4. Catalyst evaluation
Fig. 10 shows the degradation of 2,4-dichlorophenoxyacetic
acid (2,4-D) in conventional and catalytic ozonation with TiO2 and
Ni/TiO2 catalysts at pH 2.5. It is clear that the degradation pro-
files for simple and catalytic ozonation with TiO2 were similar,
reaching a 50% conversion of the pesticide in 5 min. The activity
of the Ni/TiO2 catalyst was slightly higher and the time required to
obtain a 50% conversion of 2,4-D was lower (3.8 min). Obviously,
the higher initial reaction rate with the Ni/TiO2 catalyst indicates
that the presence of NiO/Ni active sites on the surface of the TiO2
can improve the formation of OH radicals used by the degradation
of 2,4-D (see attached plot in Fig. 10). Similar results have been
reported earlier by using manganese oxides supported on meso-
porous zirconia as catalyst and 2,4-D as a model compound [48].
To be more precise, the catalytic ozonation of organic compounds
carried out in presence of supported Ni species promotes the ozone
decomposition producing a high amount of OH radicals useful for
the degradation reactions [7,8]. Therefore, if we assume that our
catalytic system Ni/TiO2 contains a surface layer of NiO and metallic
bulk nickel converting at short reaction times (10 min) almost 88%
of 2,4-D, then, these two phases (NiO/Ni) could play an important
role in the OH radical formation. In particular, our results indicated
that catalytic ozonation effectively decomposes O3 into OH, which
confirms what has been reported before [49,50]. Hence, a possible
reaction mechanism for ozone decomposition should include a first
step for ozone adsorption on the surface of the catalyst and a sec-
ond one consisting in desorption of the intermediates compounds
formed.
Fig. 8. TEM micrographs of Ni/TiO2 catalyst prepared by photocatalytic method: (a)
and (b) low magnification images; (c) high resolution.
region is composed by a peak at 531.0 eV for O2− species ascribed
to TiO2. The remaining fraction of the peak is fitted by wide peaks
at 532.6 and 533.9 eV, FWHM 3.0 and 2.5, respectively. These posi-
tions correspond to O- (OH−) and H2O adsorbed on the surface,
which is typical for titanium oxides [43]. It is clear that there is a
With regard to the intermediates formed after conventional
and catalytic ozonation with Ni/TiO2 catalyst, we found 2,4-
dichlorophenol (2,4-DCP), glycolic, glyoxylic, maleic, fumaric and
oxalic acids. The evolution of the concentration of these compounds
(not shown here) during simple or catalytic ozonation was the

J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
35
b C 1s
a Ti 2p
459.5.0 eV
284.5 eV
286.0 eV
288.4 eV
474 472 470 468 466 464 462 460 458 456
Binding Energy, eV
294 292 290 288 286 284 282 280 278 276
Binding Energy, eV
873.5 eV
855.8 eV
c O 1s
d Ni 2p
852.0 eV
531.0 eV
532.6 eV
881.0 eV
861.5 eV
533.9 eV
920 910 900 890 880 870 860 850 840 830
Binding Energy, eV
544 542 540 538 536 534 532 530 528 526
Binding Energy, eV
Fig. 9. High resolution XPS spectra of a Ni/TiO2 sample obtained by photocatalytic process: (a) Ti 2p, (b) C 1s, (c) O 1s and (d) Ni 2p.
same, however the yield of the intermediates by the second route
was lower compared with ozonation alone [51]. These results indi-
cated that ozonation and catalytic ozonation had a similar reaction
pathway. Therefore, the degradation process of 2,4-D is initiated
by the attack of ozone and/or OH radicals on the C(1)–O bond
2,4-D on the aromatic ring to give 2,4-DCP and glycolic acid, as
shown in Fig. 11. Particularly, no other aromatic compounds dif-
ferent than 2,4-dichlorophenol were detected by ozonation and
Fig. 10. Decomposition dynamics of 2,4-D in the conventional and catalytic
ozonation processes: [O3] = 25 mg L−1, catalyst dose = 0.1 g L−1, [2,4-D] = 80 mg L−1
pH = 2.5. Insert initial rate reaction as a function of method used in the 2,4-D elimi-
nation.
,
Fig. 11. Suggested reaction pathway for 2,4-D by O3 and ozonation catalyzed with
TiO2 and Ni/TiO2 at pH 2.6.

36
J.L. Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 353–354 (2012) 29–36
catalytic ozonation. However, if there their decomposition can pro-
duce maleic and fumaric acids, which are oxidized to oxalic acid.
Alternatively, oxalic acid can also be produced from glycolic acid
following a parallel path. Considering that the final byproduct of the
whole process was oxalic acid by the two routes, a complete min-
eralization of 2,4-D did not occur. Nevertheless, due to the amount
of intermediate compounds was lower by the catalytic process, it
is clear that the partial mineralization of 2,4-D was improved in
comparison with the ozonation alone.
[9] C. Bradu, L. Frunza, N. Mihalche, S.M. Avramescu, M. Neat¸a˘, I. Udrea, Appl. Catal.
B: Environ. 96 (2010) 548–556.
[10] L. Zhao, Z. Sun, J. Ma, Environ. Sci. Technol. 43 (2009) 4157–4163, and references
therein.
[11] T. Zhang, J. Ma, J. Mol. Catal. A: Chem. 279 (2008) 82–89.
[12] J.F. Lambert, M. Che, J. Mol. Catal. A: Chem. 162 (2000) 5–18.
[13] N. Toshima, T. Yonezawa, New J. Chem. 22 (1998) 1179–1201.
[14] A. Roucoux, J. Schulz, H. Patin, Chem. Rev. 102 (2002) 3757–3778.
[15] B.L. Cushing, V.L. Kolesnichenko, C.J. O’Connor, Chem. Rev. 104 (2004)
3893–3946.
[16] J.A. Dahl, B.L.S. Maddux, J.E. Hutchison, Chem. Rev. 107 (2007) 2228–2269.
[17] B. Ruiz-Camacho, M.A. Valenzuela1, J.A. Perez-Galindo, F. Pola, M. Miki-Yoshida,
N. Alonso-Vante, R.G. Gonzalez-Huerta, J. New Mater. Electrochem. Systems 13
(2010) 183–189.
4. Conclusion
[18] B. Ruiz-Camacho, R.G. Gonzalez-Huerta, M.A. Valenzuela, N. Alonso-Vante, Top.
Catal. 54 (2011) 512–518.
We developed a procedure to obtain TiO2 supported-nickel
nanoparticles by a photocatalytic deposition route. The Ni precur-
sor was easily reduced at room temperature by the photocatalytic
route in comparison with a photochemical method in presence of
different sensibilizers. Almost 90% of the Ni precursor was obtained
at short reaction time (1.5 h). The photocatalytic deposition was
dependent on the light intensity and TiO2 loading, best results
were found at medium intensities (40–60 mW cm−2) and mass of
catalyst at about 1 g L−1. It seemed that these experimental con-
ditions allowed optimal photon absorption avoiding electron-hole
recombination and agglomeration or sedimentation of TiO2.
Nickel nanoparticles ranging from 7 to 15 nm (TEM results) were
deposited on TiO2 without any thermal treatment (i.e. hydrogen
reduction) which is advantageous compared with the conventional
methods. XPS and XRD results evidenced the presence of metallic
Ni nanoparticles and nickel oxide. Ni nanoparticles were randomly
distributed on the support an easily reoxidized to NiO after stay in
contact with atmospheric air. There was no difference in the activity
results of ozonation alone compared with catalytic ozonation with
TiO2. However, the presence of two phases (NiO/Ni) in the Ni/TiO2
improved slightly the conversion of 2,4-D compared with the other
routes. It was proposed a reaction path including all the intermedi-
ates detected which were formed during the attack of both ozone
and OH radicals. It seemed that a higher partial mineralization of
2,4-D was reaching by using catalytic ozonation with the Ni/TiO2
catalyst.
[19] A. Peled, N. Mirchin, Photo-Excited Processes Diagnostics and Applications,
Kluwer Academic Publishers, Netherlands, 2003, 251–280.
[20] C. Crisafulli, S. Sciré, S. Giuffrida, G. Ventimiglia, R. Lo Nigro, Appl. Catal. A: Gen.
306 (2006) 51–57.
[21] S. Giuffrida, F.F. Condorelli, L.L. Costanzao, G. Ventimiglia, J. Nanopart. Res. 19
(2007) 611–619.
[22] S. Scirè, C. Crisafulli, S. Giuffrida, G. Ventimiglia, C. Bongiorno, C. Spinella, J. Mol.
Catal. A: Chem. 333 (2010) 100–108.
[23] J.C. Scaiano, P. Billone, C.M. Gonzalez, L. Maretti, M:l. Marin, K.L. McGilvray, N.
Yuan, Pure Appl. Chem. 81 (2009) 635–647.
[24] M. Sakamoto, M. Fujistuka, T. Majima, J. Photochem. Photobiol. C. Photochem.
Rev. 10 (2009) 33–56.
[25] M. Cerro-Alarcón, B. Bachiller-Baeza, A. Guerrero-Ruiz, I. Rodríguez-Ramos, J.
Mol. Catal. A: Chem. 258 (2006) 221–230.
[26] B. Zapata, M.A. Valenzuela, J. Palacios, E. Torres-Garcia, Int. J. Hydrogen Energy
35 (2010) 12091–12097.
[27] F. Forouzan, C.T. Richard, J.A. Bard, J. Phys. Chem. 100 (1996) 18123–18127.
[28] V. Iliev, D. Tomova, L. Bilyarska, G. Tyuliev, J. Mol. Catal. A: Chem. 263 (2007)
32–38.
[29] A. Mills, M.A. Valenzuela, J. Photochem. Photobiol. A. Chem. 165 (2004) 25–34.
[30] M. Alvarez Lemus, T. López, S. Recillas, D.M. Frías, M. Montes, J.J. Delgado, M.A.
Centeno, J.A. Odriozola, J. Mol. Catal. A: Chem. 281 (2008) 107–112.
[31] S. Velu, S.K. Gangwal, Solid State Ionics 177 (2006) 803–811.
[32] I.K. Konstantinou, T.A. Albanis, Appl. Catal. B: Environ. 49 (2004) 1–14.
[33] Y.L. Chow, G.E. Buono–Core, C.W.B. Lee, J.C. Scaiano, J. Am. Chem. Soc. 108 (1986)
7620–7627.
[34] T. Sato, H. Onaka, Y. Yonezawa, J. Photochem. Photobiol. A: Chem. 127 (1999)
83–87.
[35] Y. Yonezawa, T. Sato, S. Kuroda, J. Chem. Soc. Faraday Trans. 87 (1991)
1905–1910.
[36] I. Preda, L. Soriano, L. Alvarez, J. Mendez, F. Yubero, A. Gutierrez, J.M. Sanz, Surf.
Interf. Anal. 42 (2010) 869–873.
[37] S. Scirè, S. Giuffrida, C. Crisafulli, P.M. Riccobene, A. Pistone, J. Nanopart. Res. 13
(2011) 3217–3228.
[38] H.S. Nalwa (Ed.), Handbook of Photochemistry and Photobiology, V-1, Inorganic
Photochemistry, American Scientific Publishers, 2003.
[39] S.M. Avramescu, C. Bradu, I. Udrea, N. Mihalache, F. Rurta, Catal. Commun. 9
(2008) 2386–2391.
Acknowledgements
The authors are grateful for the financial support provided by
CONACYT (Project Nos. 83275, 49367 and 153356) and PAPIIT
IN114209.
[40] J. Li, Y. Lu, Q. Ye, M. Cinke, J. Han, M. Meyyappan, Nano Lett.
929–933.
3 (2003)
[41] S. Suzuki, Y. Watanabe, T. Ogino, S. Heun, L. Gregoratti, A. Barinov, B. Kaulich, M.
Kiskinova, W. Zhu, C. Bower, O. Zhou, Phys. Rev. B 66 (2002) 035414–035418.
[42] H. Ago, T. Kugler, F. Cacialli, W.R. Salaneck, M.S.P. Shaffer, A.H. Windle, R.H.
Friend, J. Phys. Chem. B 103 (1999) 8116–8121.
[43] N. Kruse, S. Chenakin, Appl. Catal. A: Gen. 391 (2011) 367–376.
[44] L. Soriano, I. Preda, A. Gutiérrez, S. Palacín, M. Abbate, A. Vollmer, Phys. Rev. B
75 (2007) 233417–233420.
[45] A.R. González, J.P. Holgado, R. Álvarez, G. Munuera, J. Phys. Chem. 96 (1992)
3080–3086.
[46] N.V. Smirnova, T.B. Boitsova, V.V. Gorbunova, L.V. Alekseeva, V.A. Pronin, G.S.
Konuhov, Thin Solid Films 513 (2006) 25–30.
[47] J.J. Zou, C.J. Liu, Y.P. Zhang, Langmuir 22 (2006) 2334–2339.
[48] S. Xing, C. Hu, J. Qu, H. He, M. Yang, Environ. Sci. Technol. 42 (2008) 3363–3368.
[49] C. Hu, S. Xing, J. Qu, H. He, J. Phys. Chem. C 112 (2008) 5978–5983.
[50] R.R. Giri, H. Ozaki, R. Takanami, S. Taniguchi, Water Sci. Technol. 58 (2008)
207–216.
References
[1] B. Legube, N. Karpel VelLeitner, Catal. Today 53 (1999) 61–72.
[2] B.K. Hordern, M. Ziółek, J. Nawrocki, Appl. Catal. B: Environ. 46 (2003) 639–669.
[3] J. Nawrockia, B.K. Hordern, Appl. Catal. B: Environ. 99 (2010) 27–42, and refer-
ences therein.
[4] F.J. Beltrán, F.J. Rivas, R. Montero-de-Espinosa, Appl. Catal. B: Environ. 47 (2004)
101–109.
[5] Y. Yang, J. Ma, Q. Qin, X. Zhai, J. Mol. Catal. A: Chem. 267 (2007) 41–48.
[6] F.J. Beltrán, F.J. Rivas, R. Montero-de-Espinosa, Appl. Catal. B: Environ. 39 (2002)
221–231.
[7] X. Zhang, X. Li, W. Qin, Chem. Phys. Lett. 479 (2009) 310–315.
[8] S.M. Avramescu, C. Bradu, I. Udrea, N. Mihalache, F. Ruta, Catal. Commun. 9
(2008) 2386–2391.
[51] E. Brillas, J.C. Calpe, P.L. Cabot, Appl. Catal. B: Environ. 46 (2003) 381–391.