Journal of Colloid and Interface Science 323 (2008) 182–186

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Preparation and photocatalytic properties of Fe3+ -doped Ag@TiO2 core–shell

nanoparticles
Wenjiao Wang a , Jinlong Zhang a,∗ , Feng Chen a , Dannong He b , Masakazu Anpo c
a
Lab for Advanced Materials and Institute of Fine Chemicals, East China University of Science and Technology, 130 Meilong Road, Shanghai 200237, People’s Republic of China
b
Shanghai National Engineering Research Center for Nanotechnology, 245# Jiachuan Road, Shanghai 200237, People’s Republic of China
c
Department of Applied Chemistry, Graduate School of Engineering, Osaka Prefecture University, 1-1 Gakuen-cho, Sakai, Osaka 599-8531, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Ag@TiO2 core–shell-type nanophotocatalysts have been prepared using a simple and convenient method.
Received 3 February 2008 The products were characterized by TEM, XRD, and UV–vis spectra. To make the catalysts achieve the
Accepted 27 March 2008 highest photocatalytic activity under UV light illumination, the Ag content of Ag@TiO2 was optimized.
Available online 29 April 2008
The results showed that Ag@TiO2 -doped Fe3+ extend their absorption into the visible region. Among
Keywords:
the Fe3+ -doped samples, Ag@Fe–TiO2 with low Ag content showed higher photocatalytic activity under
Titanium dioxide visible light illumination. An excessive added amount of Ag would reduce Fe3+ to Fe2+ and make them
Photodegradation difficult to be incorporated into the lattice of titania. From the experiments, we found that Fe3+ ions
Ag@TiO2 could stabilize the Ag@TiO2 colloid by holding back the aggregation of the core–shell nanoparticles.
Fe3+ doping © 2008 Elsevier Inc. All rights reserved.
Core–shell structure

1. Introduction tion of electron–hole pairs, and Fe3+ can improve the visible light
activity of titania. Due to the combined effect of Fe3+ and Ag,
There has been intense study of TiO2 as a photocatalyst because the photocatalytic activity of this core–shell-type Ag@Fe–TiO2 is
of its high chemical stability, nontoxicity, low cost, and excellent tremendously enhanced under visible light illumination.
degradation for organic pollutants [1–5]. However, the large band
gap and low quantum yields of TiO2 prevent it from practical ap- 2. Materials and methods
plications. Noble metals such as Ag, Au, Pt, and Pd deposited on
a TiO2 surface could enhance the photocatalytic efficiencies be- 2.1. Preparation of Ag@TiO2 photocatalysts
cause they act as an electron trap promoting interfacial charge
transfer processes in the composite systems [6–9]. But this type
Core–shell-type Ag@TiO2 and Ag@Fe–TiO2 nanoparticles were
of catalyst structure, though effective, results in exposing noble
metal to reactants and the surrounding medium [10]. Metals on prepared as reported in the literature [19]. This synthesis method
the surface of the semiconductor will be easily corroded and dis- includes three steps as follows:
solved. An efficient way to overcome these drawbacks is to exploit
a core–shell-type structure in which the noble metal particles are (1) Preparation of Ag clusters: Three milliliters of 1 mM aque-
introduced as cores, and the semiconductor dioxide such as TiO2 ous hydrazine solution was added into 120 mL of 1 mM
and SiO2 acted as shells [11–14]. cetyltrimethylammonium bromide (CTAB) aqueous solution,
In order to overcome the narrow absorption range of TiO2 , and the mixture was stirred for 1 min at room temperature.
many groups have been involved in doping transition metal ions After 3 mL of 50 mM aqueous AgNO3 solution was added to
into TiO2 to extend the absorption spectra into the visible re- the above mixture, the resulting solution became dark brown,
gion, Fe3+ among which has been the most extensively examined which means that Ag clusters were formed in a short time.
[15–18]. Enlightened by the respective advantages of doping and The final composite was stirred for another 10 min to make
core–shell pattern deposition, we attempt to prepare a new type this redox reaction complete.
of photocatalyst in which Ag cores are encapsulated by Fe3+ -doped (2) Preparation of Ag@TiO2 core–shell-type nanoparticles: An ad-
TiO2 shells. The Ag cores could effectively hinder the recombina- equate amount of 1 mM titanium tetraisopropoxide (TTIP)
ethanol solution was added to the highly dispersed Ag solution
under vigorously stirring for 10 min at ambient temperature to
* Corresponding author. Fax: +86 21 64252062. obtain a suspension of Ag@TiO2 core–shell-type nanoparticles
E-mail address: jlzhang@ecust.edu.cn (J. Zhang). with different Ag contents.

0021-9797/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2008.03.043
 W. Wang et al. / Journal of Colloid and Interface Science 323 (2008) 182–186 183

Fig. 1. TEM images of Ag@TiO2 nanoparticles: (a) Ag particles and (b) 80% Ag@TiO2 and (c) 60% Ag@TiO2 . Percentage of Ag = weight of Ag/weight of TiO2 .

(3) Crystallization of Ag@TiO2 nanoparticles: The excess solvent

(ethanol and water) in the Ag@TiO2 suspension was removed
by rotatory evaporation until the Ag@TiO2 powder was pro-
duced. The brown powder was hydrothermally crystallized at
180 ◦ C for 12 h. The powder was rinsed by distilled water re-
peatedly and then dried at 80 ◦ C.
(4) Preparation and crystallization of Ag@Fe–TiO2 nanoparticles:
A certain amount of Fe(NO3 )3 was added into TTIP ethanol
solution and the same method was used as illustrated above
to prepare Fe3+ -doped core–shell-type Ag@Fe–TiO2 nanoparti-
cles.

2.2. Characterization of photocatalysts

The TEM image was taken with a JEM-2010 microscope oper-

ated at 200 kV. For TEM measurements, the sample was prepared
by dispersing the colloids in ethanol and dried on a holey carbon
Fig. 2. UV–vis spectra of Ag nanoparticles and Ag@TiO2 nanoparticles with different
film. The UV–vis absorption spectra were recorded with a Varian Ag contents in solution.
Cary 500. XRD analysis of the prepared photocatalysts was car-
ried out at room temperature with a Rigaku D/max 2550 VB/PC
nanoparticles could adsorb a certain amount of residual Ag+ ions
apparatus using CuK α1 radiation (λ = 1.5406 Å) and a graphite
in the solution to carry positive charges. These cationic Ag cores
monochromator, operated at 40 kV and 30 mA.
would repulse each other to restrain the growth of Ag nanopar-
ticles effectively. So they could highly disperse and had uniform
2.3. Measurements of photocatalytic activities
morphology (Fig. 1a). After the addition of TTIP, the cationic Ag
The photocatalytic activity of each sample was evaluated in aggregation could attract the hydrolyzed Ti species by the electro-
terms of the degradation of methyl orange (MO). MO was se- static force to form a well-dispersed core–shell structure (Fig. 1b).
lected as a model pollutant because it is a common contaminant In these Ag@TiO2 composite nanoparticles, the diameters of Ag
in industrial wastewater and it has good resistance toward light cores are about 15 nm. The thickness of TiO2 shells are about
degradation. The amount of 0.5 g photocatalyst was added into a 10 nm. It is well known that an excessive amount of Ag will reduce
70-mL quartz photoreactor containing 50 mL of a 10 mg L−1 MO the photocatalytic activity of TiO2 . The followed photocatalysis test
solution. The mixture was sonicated for 10 min and stirred for has proved that the photocatalysts with lower Ag content could
30 min in the dark in order to reach the adsorption–desorption achieve higher photocatalytic activity. So it is necessary to increase
equilibrium. A 1000-W tungsten halogen lamp equipped with a the added amount of TiO2 . But the high proportion of TiO2 will
UV cutoff filter (λ > 420 nm) was used as a visible light source cause the adhesion of core–shell particles which can be observed
and a 300-W high-pressure Hg lamp for which the strongest emis- in Fig. 1c.
sion wavelength is 365 nm was used as a UV light source. The UV–vis absorption spectra of Ag colloid and Ag@TiO2 colloid
lamp was cooled with flowing water in a quartz cylindrical jacket with different Ag contents are shown in Fig. 2. The result indicates
around the lamp, and ambient temperature was maintained during that the absorption peaks of Ag@TiO2 (at 437 nm) have a red shift
the photocatalytic reaction. At given time intervals, the analytical compared to that of Ag colloid (at 416 nm), which also can be
samples were taken from the mixture and immediately centrifuged found in other references [12]. It can be explained by the dielectric
and then filtered through a 0.22-μm Millipore filter to remove the constant of the surrounding matrix. For small metal particles, the
particles. The concentration of the filtrate was analyzed by check- position of the absorption band can be described by the following
ing the absorbance at 464 nm with a UV–vis spectrophotometer formulas [20]:
(Varian Cary 100).  
λ2peak = λ2p ε ∞ + 2εm , (1)
3. Results and discussion λ2p = 4π 2 c 2mε0 / Ne 2 , (2)

The Ag@TiO2 core–shell structure was verified by TEM images where λ2p is the bulk plasma wavelength in terms of the electron
(Fig. 1). During the experiment, Ag+ ions were reduced by hy- mass, vacuum permittivity ε0 , the electron charge e, and the elec-
drazine monohydrate to form Ag nanoparticles. The yielded Ag tron density N; ε ∞ is the high-frequency dielectric constant of
184 W. Wang et al. / Journal of Colloid and Interface Science 323 (2008) 182–186

Fig. 5. Photocatalytic degradation rate of MO under UV light illumination over

Fig. 3. Wide-angle X-ray diffraction patterns of (a) Ag@TiO2 , (b) 2% Ag@TiO2 , (c) 5% Ag@TiO2 and TiO2 nanoparticles.
Ag@TiO2 , and (d) 10% Ag@TiO2 . Percentage of Ag = weight of Ag/weight of TiO2 .
“A” refers to the crystal form anatase of titanium dioxide.

Fig. 6. UV–vis spectra of Fe(NO3 )3 ethanol solution.

Fig. 4. Wide-angle X-ray diffraction patterns of (a) 60% Ag@TiO2 , and (b) 80% nation (6 h). The content of MO is 10 mg L−1 (pH 4). It is found
Ag@TiO2 . Percentage of Ag = weight of Ag/weight of TiO2 . “A” refers to the crystal that 5 wt% is the optimal content of Ag to achieve the highest
form anatase of titanium dioxide. photocatalytic activity, and any higher or lower content could both
result in lower photocatalytic activity. Among these samples, most
silver (about 4.9 ± 0.3), and εm is numerically equal to the square catalysts have a higher photodegradation rate of MO than that
of the solvent refractive index, which for water is 1.78. As known, of pure TiO2 . It can be contributed to a Schottky energy barrier
the introduction of a TiO2 shell would reduce the surface charge formed by the combination of Ag clusters and TiO2 shells. The in-
density N on the Ag core. The decrease of N will lead to the in- terface can attract light-induced electrons from the semiconductor
crease of λ2p in formula (2). It can be found in formula (1) that the TiO2 to reduce electron–hole recombination. With the increasing
value of λpeak increased following the enhancement of λp . So the content of Ag, some Ag clusters were oxidated to AgO, which can
peak of the Ag core coated with TiO2 will show a red shift com- be verified by the appearance of AgO diffraction peaks in Fig. 4.
pared with those of uncoated samples. Otherwise the intensity of The AgO cannot serve as a Schottky energy barrier. On the other
the absorption band of the Ag@TiO2 colloid reduces gradually with hand, the decreasing content of TiO2 means that fewer electrons
the decreasing proportion of Ag. were generated to take part in the photodegradation of organic
The wide-angle X-ray diffraction patterns of Ag@TiO2 and TiO2 pollutants. From the results of photodegradation, the least Ag con-
nanoparticles are shown in Fig. 3. All samples consisted of anatase tent sample does not have the best photocatalytic activity. This is
as the unique phase of titania. Samples (b), (c), and (d) show re- because the thicker TiO2 shell inhibits light-induced electrons from
flections at 38.1◦ (111), 44.3◦ (200), 64.4◦ (220), 77.4◦ (311). The arriving at the interface between the Ag cores and the TiO2 shells.
2θ values correspond to the 111, 200, 220, and 311 crystal plane Fig. 6 shows the UV–vis spectra of Fe(NO3 )3 ethanol solution,
of Ag0 , respectively. The intensity of Ag diffraction peaks increase and it can be seen that Fe(NO3 )3 has absorption bands at 220, 254,
with the increasing content of Ag. In the low Ag content sam- and 336 nm in the UV range. Among these peaks, only the absorp-
ples, the AgO diffraction peak did not appear in the XRD patterns, tion at 336 nm is not overlapped by peaks of Ag or TiO2 . In the
indicating that the Ag cores mainly exist as Ag0 in these photocat- spectra of Fe3+ -doped Ag@TiO2 (Fig. 7), the peaks of Fe(NO3 )3 at
alysts. If the Ag content is elevated to 60%, there appears an AgO 336 nm did not appear, and the positions or shapes of Ag peaks are
diffraction peak at 30.9◦ (Fig. 4). Based on the followed photocatal- not changed compared to those of undoped samples (Fig. 2). This
ysis test, we found that these Ag oxides are harmful to the activity phenomenon indicates that the Fe3+ has been incorporated into
of the catalysts. the lattice of TiO2 instead of being dissolved in solution. Also, the
Fig. 5 shows the photodegradation rate of methyl orange over intensities of Ag peaks decrease by the increase of the thickness
Ag@TiO2 nanopowders with various Ag contents under UV illumi- of the TiO2 shell. According to the results of the experiment, the
 W. Wang et al. / Journal of Colloid and Interface Science 323 (2008) 182–186 185

Fig. 7. UV–vis spectra of Ag nanoparticles and Ag@0.5% Fe–TiO2 nanoparticles with Fig. 8. Wide-angle X-ray diffraction patterns of (a) 0.5% Fe–TiO2 , (b) 1% Ag@0.5%
different Ag contents in solution. Percentage of Ag = weight of Ag/weight of TiO2 . Fe–TiO2 , (c) 2% Ag@0.5% Fe–TiO2 , (d) 5% Ag@0.5% Fe–TiO2 , and (e) 10% Ag@0.5%
Percentage of Fe3+ = mole of Fe3+ /mole of TiO2 . Fe–TiO2 . “A” refers to the crystal form anatase of titanium dioxide.

surface plasmon absorption of Fe3+ -doped Ag@TiO2 (at 454 nm)

also had a red shift than Ag suspensions (at 416 nm). Compared
to the peaks of undoped samples (at 437 nm in Fig. 2), the red
shift of Fe3+ -doped Ag@TiO2 is extended. It is because the Fe3+
had a substitute of Ti4+ in the TiO2 lattice to enhance the electron
negativity of TiO2 shells. The charged TiO2 shells have a rejection
force to negative charges, so the electron attraction of the Ag core
had been weakened. Thus the electron density N of Ag cores has
been decreased. Reduction of N will lead to the increase of λ p
(formula (1)), and to the increase of λpeak finally (formula (2)).
Another interesting discovery is that the Fe3+ -doped in the lat-
tice of TiO2 shells could enhance the stability of core–shell-type
colloids under certain conditions. The stability of Ag@TiO2 and
Ag@Fe–TiO2 colloids has also been investigated. We found that un-
der normal conditions (without adding a donor), the Fe3+ -doped
Fig. 9. Photocatalytic degradation rate of MO under visible light (λ > 420 nm) illu-
Ag@TiO2 colloids were not as stable as the undoped samples. For mination over Ag@0.5% Fe–TiO2 nanoparticles.
example, 80% Ag@TiO2 colloid will be stable in 2 h, but for 80%
Ag@0.5% Fe–TiO2 , the stable time was shortened to 1.5 h. Fe3+
doped in the lattice of TiO2 would enhance the electron negativ-
ity of TiO2 shells. Thus fewer electrons are attracted to the surface
of Ag core, which causes the repulsion force between particles to
weaken. So the aggregation of Ag makes the Ag@0.5% Fe–TiO2 se-
ries unstable. So the Fe3+ -doped samples are less stable than the
Ag@TiO2 series without donors. But by adding a donor, the stability
of Ag@Fe–TiO2 will be greatly enhanced and surpass the Ag@TiO2
series. In order to avoid introducing other impurities, we choose
N2 H4 ·H2 O as a donor. For example, 80% Ag@TiO2 colloid will be
stable in 2 h; but for 80% Ag@0.5% Fe–TiO2 , the stable time has
been shortened to 1.5 h. If we add a little N2 H4 ·H2 O into the col-
loids, the stable time of 80% Ag@TiO2 is about 19 h; but for the
80% Ag@0.5% Fe–TiO2 , it could remain stable even after 4 days.
This rule is also suitable for other samples. After we add a donor,
both Ag cores and doped Fe3+ could attract electrons. As a re-
sult, the whole core–shell particle will carry more charges to avoid
aggregating with others. In this case, Fe3+ -doped samples will be
more stable than undoped products.
The wide-angle X-ray diffraction patterns of 0.5% Fe3+ -doped
Ag@TiO2 and TiO2 nanoparticles are shown in Fig. 8. The crystal Fig. 10. Proposed mechanism for the synergistic effects of Fe3+ and Ag.
forms of TiO2 in all samples are anatase. No appearance of an iron
oxide characteristic peak indicates that Fe3+ exists in the crystal Fig. 9 shows the photodegradation rate of methyl orange over
lattices of TiO2 . But the characteristic diffraction peaks of AgO ap- Ag@0.5% Fe–TiO2 nanopowders with various Ag contents under
pear in most of the samples except the product with 1 wt% Ag visible light illumination (3 h). The content of MO is 10 mg L−1
content. The formation of AgO is promoted by the addition of Fe3+ . (pH 4). The amount of doped Fe3+ is 0.5 mol% in all these sam-
When the Ti source was added into the Ag suspension, the Fe3+ in ples, as this is an accepted optimal doped proportion of iron ions.
the TTIP ethanol solution would snatch electrons from Ag0 , so a The sample with 1.0 wt% Ag achieved the best photocatalytic ac-
part of Ag0 had turned to AgO. tivity and it is the only sample which has a quicker degradation
186 W. Wang et al. / Journal of Colloid and Interface Science 323 (2008) 182–186

rate than the sample without Ag cores. The synergistic effects of prevent Fe3+ from being doped into the lattice of TiO2 and lead to
Fe3+ and Ag on raising the photocatalysis are shown in Fig. 10. At the oxidation of partial Ag. AgO does not exist in the sample with
this ratio of Fe3+ and Ag, the Fe3+ could successfully substitute 1 wt% Ag@0.5% Fe–TiO2 , and this product does best in a photo-
Ti4+ in the TiO2 lattice to introduce a dopant energy level into catalysis process under visible light illumination. Besides, the Fe3+
the band gap of TiO2 . As a result, the light absorption of TiO2 can doped in Ag@TiO2 will stabilize the colloid of core–shell nanopar-
be extended into the visible region. Under the illumination of visi- ticles at the existence of a donor.
ble light, the electrons can be excited from the dopant level to the
conduction band. These light-induced electrons penetrated through Acknowledgments
TiO2 shell to the Ag0 core so as to enhance the electron–hole sep-
aration. But the photocatalytic activity of 2, 5, and 10% Ag@0.5% This work has been supported by Shanghai Nanotechnology
Fe–TiO2 is lower than that of 0% Ag@0.5% Fe–TiO2 . There are two Promotion Centre (0652nm045, 0752nm001), Science and Technol-
main reasons. On one hand, increasing the content of Ag would ogy Commission of Shanghai Municipality (07JC14015), National Na-
lead to the transformation from Ag0 to AgO, which has been illus- ture Science Foundation of China (20577009, 20773039), National
trated in Fig. 8. At the same time, some Fe3+ had been reduced Basic Research Program of China (973 Program, 2007CB613306,
into Fe2+ during the preparation procedure. The electron transfer 2004CB719500), and the Ministry of Science and Technology of
process can be expressed in the following formula: China (2006AA06Z379, 2006DFA52710). A part of research work
Ag0 + nFe3+ → Agn+ + nFe2+ . (3) was finished in Shanghai Nanotechnology Joint Lab.
4+ 3+ 2+
The semidiameters of Ti , Fe , and Fe are 0.61, 0.64 and References
0.76 Å, respectively. From the data we can find that Fe3+ could
easily get into the crystal lattice instead of Ti4+ due to the close [1] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C 1 (2000) 1.
semidiameters of the two metal ions. But the size of Fe2+ is much [2] C. Hu, J.C. Yu, Z. Hao, P.K. Wong, Appl. Catal. B 42 (2003) 47.
bigger than Ti4+ so that it is hard to be incorporated into the lat- [3] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995)
69.
tice of the Ag@TiO2 composite.
[4] J.C. Yu, W. Ho, J. Yu, H. Yip, P.K. Wong, J. Zhao, Environ. Sci. Technol. 39 (2005)
In order to examine the effect of Fe3+ on the photocatalytic ac- 1175.
tivity, we have fixed the Ag amount at 1 wt%. Because the 1 wt% [5] P.V. Kamat, Chem. Rev. 93 (1993) 267.
Ag content sample has the best photocatalysis under UV illumi- [6] D. Behar, J. Rabani, J. Phys. Chem. B 110 (2006) 8750.
nation. The proportion of Fe3+ is varied from 0.5 to 0.25 mol%. [7] S.K. Kim, S.J. Hwang, W. Choi, J. Phys. Chem. B 109 (2005) 24260.
[8] X.Z. Li, F.B. Li, Environ. Sci. Technol. 35 (2001) 2381.
After 3 h visible light illumination, the decomposition rates of 1%
[9] X. You, F. Chen, J. Zhang, M. Anpo, Catal. Lett. 102 (2005) 247.
Ag@0.5% Fe–TiO2 and 1% Ag@0.25% Fe–TiO2 are 56.3 and 56.7%, [10] T. Hirakawa, P.V. Kamat, J. Am. Chem. Soc. 127 (2005) 3928.
respectively. This result indicates that the variation of the Fe3+ [11] R.T. Tom, A.S. Nair, N. Singh, M. Aslam, C.L. Nagendra, R. Philip, K. Vijayamo-
content does not affect the photocatalytic activity greatly. The Ag hanan, T. Pradeep, Langmuir 19 (2003) 3439.
content in the Ag@0.5% Fe–TiO2 composite plays an important role [12] T. Ung, L.M. Liz-Marzán, P. Mulvaney, Langmuir 14 (1998) 3740.
[13] I. Pastoriza-Santos, D.S. Koktysh, A.A. Mamedov, M. Giersig, N.A. Kotov, L.M.
on the photocatalysis.
Liz-Marzán, Langmuir 16 (2000) 2731.
[14] S.C. Chan, M.A. Barteau, Langmuir 21 (2005) 5588.
4. Summary [15] J. Zhu, F. Chen, J. Zhang, H. Chen, M. Anpo, J. Mol. Catal. A 216 (2004) 35.
[16] W. Choi, A. Termin, M.R. Hoffmann, J. Phys. Chem. 98 (1994) 13669.
A series of Ag@TiO2 core–shell-type nanophotocatalysts have [17] K. Nagaveni, M.S. Hegde, G. Madras, J. Phys. Chem. B 108 (2004) 20204.
[18] J. Zhu, Z. Deng, F. Chen, J. Zhang, H. Chen, M. Anpo, J. Huang, L. Zhang, Appl.
been prepared by a one-pot method and 5 wt% is the optimal con-
Catal. B 62 (2006) 329.
tent of Ag to achieve the highest photocatalytic activity under UV [19] H. Sakai, T. Kanda, H. Shibata, T. Ohkubo, M. Abe, J. Am. Chem. Soc. 128 (2006)
light illumination. Fe3+ was doped into Ag@TiO2 which extends its 4944.
absorption into the visible region. But a high content of Ag will [20] T. Ung, L.M. Liz-Marzán, P. Mulvaney, J. Phys. Chem. B 103 (1999) 6770.