Journal of Hazardous Materials 246–247 (2013) 10–17

Contents lists available at SciVerse ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Arsenite removal from aqueous solutions by ␥-Fe2 O3 –TiO2 magnetic

nanoparticles through simultaneous photocatalytic oxidation and adsorption
Lian Yu, Xianjia Peng ∗ , Fan Ni, Jin Li, Dongsheng Wang, Zhaokun Luan
State key Laboratory of Environmental Aquatic Chemistry, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, P.O. Box 2871, Beijing 100085, China

h i g h l i g h t s

 Fe–Ti binary oxide nanoparticles possess the performance of ␥-Fe2 O3 and TiO2 .
 The nano-material can efficiently oxidize As(III) to As(V) under UV light.
 The nano-material showed significantly high adsorption capacity of As(V).
 The nano-material can be separated efficiently by a magnetic separation method.
 The nano-material can be recovered effectively and can be put into cyclic use.

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

Article history: A novel Fe–Ti binary oxide magnetic nanoparticles which combined the photocatalytic oxidation prop-
Received 21 June 2012 erty of TiO2 and the high adsorption capacity and magnetic property of ␥-Fe2 O3 have been synthesized
Received in revised form 15 October 2012 using a coprecipitation and simultaneous oxidation method. The as-prepared samples were character-
Accepted 2 December 2012
ized by powder XRD, TEM, TG-DTA, VSM and BET methods. Photocatalytic oxidation of arsenite, the
Available online 10 December 2012
effect of solution pH values and initial As(III) concentration on arsenite removal were investigated in
laboratory experiments. Batch experimental results showed that under UV light, As(III) can be efficiently
Keywords:
oxidized to As(V) by dissolved O2 in ␥-Fe2 O3 –TiO2 nanoparticle suspensions at various pH values. At the
␥-Fe2 O3 –TiO2
Photocatalytic oxidation
same time, As(V) was effectively removed by adsorption onto the surface of nanoparticles. The maximum
Adsorption removal capability of the nano-material for arsenite was 33.03 mg/g at pH 7.0. Among all the common
Arsenite coexisting ions investigated, phosphate was the greatest competitor with arsenic for adsorptive sites on
Magnetic separation the nano-material. Regeneration studies verified that the ␥-Fe2 O3 –TiO2 nanoparticles, which underwent
five successive adsorption–desorption processes, still retained comparable catalysis and adsorption per-
formance, indicating the excellent stability of the nanoparticles. The excellent photocatalytic oxidation
performance and high uptake capability of the magnetic nano-material make it potentially attractive
material for the removal of As(III) from water.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction as hyperkeratosis, skin lesions, cancers of brain, kidney, liver, and

Arsenic contamination in natural water is a great threat to
millions of people in many regions of the world [1,2]. As a widely
distributed element in nature, arsenic can get mobilized into
surface water/groundwater from soils and ores via both natural
process and/or human activities [3]. Countries with arsenic pol-
lution include India, Bangladesh, China, United States, Vietnam,
Chile, Mexico, Argentina, Poland, Hungary, Canada, New Zealand
and Japan [4]. Chronic exposure to arsenic through either agricul-
ture products irrigated by contaminated water or contaminated
drinking water can lead to a lot of serious health problems such
∗ Corresponding author. Tel.: +86 10 62849198; fax: +86 10 62849198.
E-mail address: xjpeng@rcees.ac.cn (X. Peng).
stomach [5]. Accordingly, the United States Environmental Protec-
tion Agency (USEPA) adopted a new maximum contaminant level
(MCL) of 10 ␮g/L for arsenic in drinking water in 2001 and required
compliance with this new level since 2006 [5]. The more stringent
drinking water standard on arsenic requires the installation of new
water treatment systems and upgrading of existing ones.
Arsenic exists in different forms, the common valencies in
ground water sources are As(III) (arsenite) and As(V) (arsenate),
and inorganic trivalent form of As(III) is the most toxic among the
various forms of arsenic present in natural water. The inorganic
hydrolysed species are present as H3 AsO3 , H2 AsO3 − , HAsO3 2− ,
AsO3 3− , H3 AsO4 , H2 AsO4 − , HAsO4 2− and AsO4 3− [6]. Several
techniques are available for the removal of arsenic from contami-
nated water sources, including adsorption/ion exchange processes,
coagulation–precipitation processes, and membrane processes [7],

0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jhazmat.2012.12.007
 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
which remove As(V) from water efficiently. As(III) is generally
reported to have low affinity to the surface of various adsorbents
compared with As(V) because As(III) exists mainly as nonionic
H3 AsO3 in natural water with pH value ranging from weakly acidic
to weakly alkaline [8–10]. To achieve higher arsenic removal, a
pretreatment for As(III) oxidation is therefore usually involved
and then As(V) is removed through adsorption, coagulation, ion
exchange or membrane processes. Many oxidants or oxidant-
generating systems have been tested to oxidize As(III) to As(V),
including oxygen and/or ozone [11], hydrogen peroxide [12],
chlorine [13], manganese dioxide[14], KMnO4 [8], UV/H2 O2 [15],
UV/Fe(III) complexes [16,17], Fe(II)/H2 O2 [18], potassium ferrate
[19], and UV/TiO2 [9,15,20–22]. As UV/TiO2 system could provide
a robust and relatively low-cost approach, without either the
addition of harmful materials or the possible production of toxic
disinfection byproduct (DBPs) or disinfection byproduct precursors
(DBPPs) [3], many efforts have been made on the photocatalytic
oxidation of As(III) with TiO2 under UV light illumination since the
first report by Rajeshwar and coworkers in 1999 [15].
Although UV/TiO2 system can oxidize As(III) to As(V), the
adsorption capacity of TiO2 is very low, resulting in the ineffi-
cient removal of As(V) [23]. This shortcoming could be overcome
by adopting a particular material with better adsorption perfor-
mance to be introduced in the system. More recently, a water
decontamination process which can harvest As-adsorbed magnetic
nanoparticles from a suspension using an external magnetic field
has also been proposed [24]. Many studies proved that maghemite
(␥-Fe2 O3 ), which is magnetic, possesses excellent adsorption prop-
erty in water treatment [24–26]. As(V) can be efficiently adsorbed
onto the surface of maghemite and the As-adsorbed adsorbent
can be easily separated from water using magnetic separation
method. As titanium dioxide can be considered as isomorphic to
maghemite (octahedral) with the same Pauli radius, ␥-Fe2 O3 could
be incorporated onto the surface of TiO2 lattice. Thus, the compos-
ite nanoparticles may possess excellent removal capability of As(III)
and can be easily separated from water using a magnetic separation
process after photocatalytic oxidation and adsorption.
On the basis of the above consideration, in this work, a novel
nanostructured bifunctional material, ␥-Fe2 O3 –TiO2 is synthe-
sized. This material combines the catalytic characteristics of TiO2
and adsorption characteristics of ␥-Fe2 O3 . The composite pos-
sesses the synergy of the photocatalytic ability of TiO2 for oxidation
of As(III) and the adsorption performance for arsenate and mag-
netism of ␥-Fe2 O3 . The main objective of this research is to: (1)
prepare ␥-Fe2 O3 –TiO2 nanoparticles using a coprecipitation and
simultaneous oxidation method for effective arsenite removal; (2)
characterize the nano-material with a variety of techniques; (3)
investigate its photocatalytic oxidation of As(III) to As(V), evaluate
its arsenic adsorption capabilities and to examine the influence of
co-existing ions on As(III) removal; and finally (4) study the regen-
eration performance of the nano-material.
2. Materials and methods
2.1. Materials
All chemicals used were of analytical grade and were pur-
chased from Beijing Chemical corporation (Beijing, China). All stock
and fresh solutions were prepared with deionized water. Stan-
dards for calibration were prepared from As(V) and As(III) standard
reference sodium (meta) arsenic. Stock solutions of As(III) were pre-
pared by dissolving sodium arsenite (NaAsO2 ) in deionized water.
Stock solution (1000 mg/L) was frozen to prevent oxidation. Arsenic
working solutions were freshly prepared by diluting arsenic solu-
tions with deionized water. The concentrations of arsenic species
11
were always given as elemental arsenic concentration in this
study.
2.2. Preparation of -Fe2 O3 –TiO2 nanoparticles
␥-Fe2 O3 –TiO2 nanoparticles were prepared by oxidation and
coprecipitation method. First, ferric chloride (FeCl3 ·6H2 O, 5.11 g)
and ferrous chloride (FeCl2 ·4H2 O, 2.21 g) were dissolved in deion-
ized water at a molar ratio of 1.7:1. The solution was kept in the dark
and stirred magnetically for 5 min, and then 2 mL of titanium tetra-
chloride (TiCl4 ) was added into the solution. NH4 OH solution (25%)
was added dropwise into the solution using a syringe pump under
vigorous stirring. After an initial brown precipitate, a black precip-
itate formed. When pH reached 9.0, NH4 OH solution was stopped
adding to the solution, the mixture was diluted into 200 mL with
deionized water and stirred vigorously for 30 min at 50 ◦ C water
bath. Then the mixture was placed on a strong magnet and the par-
ticles were then separated from the solution. After the separation,
the particles were washed 3–5 times with deionized water and then
dried at 40 ◦ C for 24 h. Finally, these particles were crushed, calci-
nated under the nitrogen atmosphere at 400 ◦ C for 1 h to obtain
crystallized anatase and calcined under the oxygen atmosphere at
250 ◦ C for 2 h. The dry material was stored in a desiccator for fur-
ther use. As control samples, ␥-Fe2 O3 and TiO2 nanoparticles were
synthesized under the same condition as the composite sample in
the absence of TiCl4 or ferric (ferrous) salt.
2.3. Characterization of the nanoparticles
X-ray diffraction (XRD) (Philips PW1830) patterns were col-
lected using a Bruker D8 Advance X-ray diffractometer with
monochromated high-intensity Cu K␣ radiation (=1.5406 Å) in
a 2 range of 10–80◦ . Morphologies of the particles were exam-
ined by the transmission electron microscopy (TEM) (JEOL-2010).
The Brunauer-Emmett-Teller surface area (SBET ) were derived from
a Micromeritics ASAP 2000 surface area analyzer. Thermogravi-
metric analysis (TGA) was performed using thermogravimetric
analyzer (PerkinElmer TGA 7; DSC 7) with a heating rate of
5 ◦ C min−1 in an air flow. The zeta potential of nanoparticles
obtained at different pH was determined using a zeta poten-
tial analyzer (Zetasizer 2000, Malvern, UK). Vibrating sample
magnetometer (VSM) (Ming seton’s company, 7037/9509-P) was
employed to determine magnetic properties of magnetic nanopar-
ticles.
2.4. Batch removal experiments
A 20 W UV lamp equipped with circulating cooling water was
placed above (10 cm) the water surface and used as light source.
The reactor was open to the ambient air so as to reach the air-
equilibrated condition. The pH value of the solution was adjusted
with HCl or NaOH solution to the desired values. The concentrations
of ␥-Fe2 O3 –TiO2 nanoparticles were fixed at 0.5 g/L in all experi-
ments. The concentrations of As(III) were fixed at 1.01 mg/L in the
effect of solution pH on As(III) removal and photocatalytic oxida-
tion experiments, in the effect of initial concentration on As(III)
removal, the initial concentrations of As(III) ranged from 0.1 to
50 mg/L. The pH of the solution was maintained at 7.0 ± 0.1 in ini-
tial concentration on As(III) removal and photocatalytic oxidation
experiments, in the effect of solution pH on As(III) removal, the
solution pH ranged from 3.0 to 9.0. During the experiment, the
arsenic solution was stirred adequately to disperse ␥-Fe2 O3 –TiO2
nanoparticles to ensure a good contact of ␥-Fe2 O3 –TiO2 nanopar-
ticles with the arsenic species. The reaction time varied from 10
to 720 min. To examine both the adsorption capability in dark and
photocatalytic oxidation effect of UV light illumination, the As(III)
12 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
Fig. 1. XRD patterns of synthesized nanoparticles.
removal experiments were carried out with and without the UV
light illumination in the presence of ␥-Fe2 O3 –TiO2 nanoparticles
respectively. For comparison, As(III) solution was also just treated
by UV light illumination without nanoparticles. The nanoparticles
were recovered by using magnetic separation method with a per-
manent magnet (2300 G), made of Nd–Fe–B. The concentration
of As (tot) and As(III) was determined using hydride generation
atomic fluorescence spectrometry (HG-AFS) (Beijing Beifenruili
Analytic Instrument Corporation, AF-610B) after 0.45 ␮m mem-
brane filtration.
3. Results and discussion
3.1. Characterization of -Fe2 O3 –TiO2 nanoparticles
3.1.1. XRD analysis
XRD was used to investigate the crystal phase structure, crystal-
lite size and crystallinity of ␥-Fe2 O3 –TiO2 , which all play important
roles in photocatalytic activity and adsorption capability. In Fig. 1,
XRD patterns of ␥-Fe2 O3 and ␥-Fe2 O3 –TiO2 are provided. With Ti-
doping, the peak intensity of ␥-Fe2 O3 slightly increased and the
width of diffraction peak became narrower. Therefore, it is reason-
able to deduce that Ti doping may result in bigger crystallite size
of ␥-Fe2 O3 –TiO2 compared with ␥-Fe2 O3 . This was confirmed by
TEM observation shown in Fig. 2. As the crystal lattice are octahe-
dral and the radii of Ti4+ (0.68 Å) and Fe3+ (0.64 Å) ions are very
similar, titanium dioxide can be considered as isomorphic to ␥-
Fe2 O3 with the very similar Pauli radius. So ␥-Fe2 O3 can locate at
interstices of TiO2 unit cell or adsorb onto the surface of TiO2 unit
cell by chemical bond or electrostatic force [27].
Fig. 2. TEM image of (a) ␥-Fe2 O3 and (b) ␥-Fe2 O3 –TiO2 nanoparticles.
As can be seen from Fig. 1, besides ␥-Fe2 O3 , a weak diffrac-
tion peak was identified which could be assigned to anatase XRD
diffraction peak. However, no rutile phase was found in the XRD
pattern. This observation suggests that titanium additive exists as
TiO2 (anatase) at a very small quantity and is not incorporated into
the ␥-Fe2 O3 structure. Since it is generally accepted that anatase
phase of titania is more efficient as photocatalyst than brookite or
rutile phase [28], the synthetized ␥-Fe2 O3 –TiO2 nanoparticles may
show high photocatalytic activity under UV light. The crystallite
size of ␥-Fe2 O3 phase is ∼31 nm, obtained from the strongest XRD
peak (at 2 ∼ 35.7◦ for ␥-Fe2 O3 phase) by Scherrer’s formula [3]:
0.9
D= (1)
ˇ cos 
where  is the average wavelength of the X-ray radiation, ˇ is the
line-width at half-maximum peak position, and  is the diffracting
angle.
3.1.2. TEM observation
The crystalline structure of the obtained ␥-Fe2 O3 –TiO2 was
characterized by TEM. Fig. 2 shows the TEM image of ␥-Fe2 O3 –TiO2
nanoparticles. For comparison, TEM images of ␥-Fe2 O3 and TiO2
nanoparticles were also provided. From Fig. 2b, it can be seen that
the particles which may contain ␥-Fe2 O3 , TiO2 and ␥-Fe2 O3 –TiO2
are nanosized with nonuniform shapes. The average particle size of
␥-Fe2 O3 –TiO2 nanoparticles is around 25–35 nm, while the aver-
age size of ␥-Fe2 O3 and TiO2 nanoparticles are around 15–20 nm,
35–50 nm, respectively as shown in Fig. 2a and c, which suggests
an obvious increase in the particle size in case of Ti-doped ␥-Fe2 O3
nanoparticles. This is consistent with XRD analysis. As doping can
reduce the crystallinity of ␥-Fe2 O3 , the nanoparticles with low
crystallinity tend to agglomeration when treated at 400 ◦ C. The
BET specific surface area of these nanoparticles is measured to
be 153.9 m2 /g, corresponding to an average particle diameter of
∼35 nm, which is close to the calculated ␥-Fe2 O3 phase crystallite
size and TEM observation.
3.1.3. Thermal analysis
Fig. 3 presents the TG/DTG profiles of ␥-Fe2 O3 –TiO2 nanopar-
ticles, which shows a total weight loss of 7.16%. The TG curve
of ␥-Fe2 O3 –TiO2 nanoparticles is divided into three stages. The
first stage with weight loss of 5.04% occurred at 30–200 ◦ C and
two endothermic peaks at ca. 53.8 ◦ C and 191.5 ◦ C are observed
in DTG curve, mainly can be due to the removal of surface-
adsorbed water and CO2 molecules. The second stage happened
at 200–500 ◦ C resulting in 1.60% of weight loss, which could be
ascribed to the removal of water of crystallization (as supported
by the DTG peak around 275 ◦ C). Also in this stage, anatase phase
 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
Fig. 3. TG/DTG profiles of ␥-Fe2 O3 –TiO2 nanoparticles.
Fig. 4. VSM hysteresis loops of ␥-Fe2 O3 and ␥-Fe2 O3 –TiO2 nanoparticles.
formed and ␥-Fe2 O3 was transformed to ␣-Fe2 O3 . The subse-
quent step of minor weight loss of 0.51% took place at 500–900 ◦ C,
possibly because of the loss of hydroxyl groups of the parti-
cle surface. In this stage, anatase can be transformed to rutile
phase.
3.1.4. VSM
Fig. 4 shows the magnetic hysteresis loops of the samples. The
maximum saturation magnetization of ␥-Fe2 O3 and ␥-Fe2 O3 –TiO2
are 58.75 emu/g, 46.28 emu/g, respectively. It indicated that all
samples had superparamagnetism. The magnetization curve and
demagnetization curve are coincidence, and no hysteresis phe-
nomenon was found. Coercivity and remanent magnetization are
Fig. 5. Effect of pH on (a) zeta potential (0.01 M KNO3 ) and (b) arsenic removal by ␥-Fe2 O3 –TiO2 and ␥-Fe2 O3 nanoparticles. Initial As(III) concentration 1.01 mg/L; adsorbent
dosage 0.5 g/L and reaction time 3 h.
13
equal to zero. But the hysteresis loops of ␥-Fe2 O3 –TiO2 are weaker
than that of ␥-Fe2 O3 , this could be main due to a lower amount of
␥-Fe2 O3 in ␥-Fe2 O3 –TiO2 composite. What’s more, its maximum
saturation magnetization, 46.28 emu/g is strong enough to main-
tain their good performance of magnetic recovery [29].
3.2. Effect of solution pH on As(III) removal
Fig. 5b shows the effect of solution pH on photocatalytic oxi-
dation of As(III) and arsenic (including As(III) and As(V)) sorption
onto the ␥-Fe2 O3 –TiO2 and ␥-Fe2 O3 nanoparticles. The experiment
was conducted under UV light, the initial As(III) concentration was
1.01 mg/L and the reaction time was 3 h. As can be seen from Fig. 5b,
the percent removal of arsenic by the ␥-Fe2 O3 increased from 35.56
to 64.38% with the increase in pH from 3.0 to 9.0, while the percent
removal of arsenic by the ␥-Fe2 O3 –TiO2 decreased from 99.82 to
91.45% with increasing solution pH. The ␥-Fe2 O3 –TiO2 nanopar-
ticles show a relatively high removal efficiency for As(III) in the
whole pH range studied. By contrast, the ␥-Fe2 O3 nanoparticles
show a relatively low removal efficiency for As(III).
The different removal efficiencies for As(III) of ␥-Fe2 O3 and ␥-
Fe2 O3 –TiO2 nanoparticles reflect the different removal mechanism
for As(III). In the process of adsorption, As(III) could be oxidized to
As(V) by dissolved O2 in ␥-Fe2 O3 –TiO2 suspension under UV light
(as can be seen in the following study). However, little As(III) could
be oxidized to As(V) in ␥-Fe2 O3 suspension under UV light. The zeta
potentials of the nano-materials as a function of solution pH are
shown in Fig. 5a. It can be seen that the zero point of zeta potential
of ␥-Fe2 O3 and ␥-Fe2 O3 –TiO2 are round 6.5 and 6.7, respectively.
As(V) can be adsorbed by the adsorbent via the electrostatic inter-
action and surface-complexation below the zeta potentials of the
adsorbent. However, the As(V) sorption on the adsorbent at pH
above zero point of zeta potential cannot be interpreted via the
electrostatic interaction since the electrostatic interaction between
the adsorbent and As(V) should be electrically repulsive. Similarly,
electrostatic attraction is impossible in As(III) adsorption on the
nanoparticles, as As(III) is present as neutral HAsO2 at pH below
8.0 and H2 AsO3 − becomes dominant at pH above 9.2 [30]. At low
pH, few hydroxyl groups were protonated, leading to a lower sorp-
tion capacity for As(III). However, the ␥-Fe2 O3 nanoparticles had
an increasing sorption capacity for As(III) as solution pH increased.
One possible reason is that the hydroxyl groups on the adsorbent
surface played a dominant role in the sorption process [31], as tem-
perate alkaline condition is favorable for surface hydroxyl groups
generating.
14 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
Fig. 6. Effect of initial As(III) concentration on (a) As(III) and (b) arsenic removal by ␥-Fe2 O3 –TiO2 nanoparticles. Dosage 0.5 g/L; solution pH 7.0 ± 0.1 and reaction time 3 h.
3.3. Effect of initial concentration on As(III) removal
Fig. 6 shows the effect of initial As(III) concentration on
As(III) and arsenic removal by ␥-Fe2 O3 –TiO2 nanoparticles. It was
observed that the percent removal of As(III) and arsenic decreased
while the As(III) and arsenic removal capability increased with the
increase in initial As(III) concentration. This decrease of percent
As(III) and arsenic removal may be attributed to the fact that at
higher As(III) concentration, the number of active sites on adsor-
bent surface is not enough to accommodate As(III) and As(V) anions,
what’s more, there are not enough active sites for the oxidation of
As(III) at higher As(III) concentration. However, at low As(III) con-
centration, the ratio of surface active sites to total As(III) is high and
therefore As(III) ions can be fully oxidated to As(V) and can interact
with the active sites on adsorbent surface sufficiently.
Fig. 6a and b demonstrates that the arsenite removal capabil-
ity and arsenic adsorption capacity of ␥-Fe2 O3 –TiO2 nanoparticles
did not reach the saturation even when the equilibrium concen-
tration of As(III) and arsenic reached 34.56 mg/L and 36.74 mg/L,
respectively. Fig. 6a and b gives the maximal As(III) removal capa-
bility of 33.03 mg/g and maximal arsenic adsorption capacity of
26.52 mg/g, respectively. These results suggest that ␥-Fe2 O3 –TiO2
magnetic nano-material is effective for the removal of As(III) from
water.
3.4. As(III) removal through simultaneous photocatalytic
oxidation and adsorption
As the photocatalytic reactions occur only at/near the photo-
catalyst surface, it is essential that photocatalysts have a good
adsorption capability to As(III) [3]. It has been demonstrated pre-
viously that ␥-Fe2 O3 could be an effective adsorbent for arsenic
species [17,19,24]. The experiments were carried out with ␥-
Fe2 O3 –TiO2 nanoparticles in dark to ensure no photocatalytic
interaction occurred. From Fig. 7a, around 56.44% As(III) in the solu-
tion was adsorbed in just 10 min, this suggest that As(III) could be
adsorbed onto ␥-Fe2 O3 –TiO2 nanoparticles in a short time.
As demonstrated in Fig. 7b, the experimental data fitted very
well into a pseudo-second-order rate kinetic model. This observa-
tion is very similar to the result reported by Meng and coworkers
on nanocrystalline TiO2 [32]. A pseudo-second-order rate expres-
sion [33] is defined by Eq. (2) and its integrated form is given in Eq.
(3):
dqt
= Kad (qe − qt )2
dt
t 1 t
= +
qt (K2 qe 2 ) qe
where qe and qt are the amount (mg/g) of arsenic adsorbed at equi-
librium and at time t, respectively, and Kad is the rate constant
of adsorption (g/(mg min)). The kinetics parameters that fitted the
experimental data of kinetic were summarized in Table 1. The appli-
cability of the pseudo-second-order rate model is quantified by the
square of the correlation coefficient r(r2 ), and the closeness of r2 to
1 indicates that the model fitted the experimental data accurately.
As the experimental conditions are different, it is almost impossi-
ble to directly compare the parameters of adsorption among the
reports in literature. In ref [6], the specific surface area of the syn-
thesized magnetic binary oxide particles is 123.38 m2 /g, which is
similar to the specific surface area of ␥-Fe2 O3 –TiO2 nanoparticles
in this study (153.9 m2 /g). The experiment was conducted without
UV light, the initial As(III) concentration of the solution is 1.01 mg/L,
the pH value is 7.0, and the amount of magnetic binary oxide par-
ticles is 1.0 g/L. The reported adsorption Kad is 0.29 g/(mg h). Their
adsorption Kad to As(III) is about twentieth of that in our exper-
iment with initial concentration of As(III) at 1.01 mg/L, pH value
at 7.0, and the amount of ␥-Fe2 O3 –TiO2 nanoparticles at 0.5 g/L.
Thus these comparisons demonstrate ␥-Fe2 O3 –TiO2 nanoparticles
possess a excellent adsorption capability of As(III).
Fig. 8 demonstrates the results of arsenic species concentrations
vs treatment time under various conditions in photocatalytic oxi-
dation and adsorption experiment. Fig. 8a shows the remaining
concentration of As(III) after three different treatments. With only
UV light illumination, no obvious decrease in the concentration of
As(III) was observed. This observation suggests that little As(III)
could be oxidized to As(V) with UV light alone. Although ␥-
Fe2 O3 –TiO2 nanoparticles possess a good adsorption capability for
As(III), the residual As(III) concentration after 3 h treatment was
still at 0.35 mg/L, which is much higher than the EPA standard
(10 ␮g/L) for drinking water. Further decrease of As(III) concen-
tration was observed when both ␥-Fe2 O3 –TiO2 nanoparticles and
UV light illumination were present. At the beginning of the reac-
tion under UV light, the adsorption of As(III) onto ␥-Fe2 O3 –TiO2
nanoparticles still played the major role. As the illumination
time increased, more and more As(III) was oxidized to As(V)
and then adsorbed by ␥-Fe2 O3 –TiO2 nanoparticles. After 12 h UV
light illumination, the residual As(III) concentration was 4.5 ␮g/L,
which is lower than the 10 ␮g/L limit and represents another
magnitude degree of decrease compared to pure adsorption
effect.
Table 1
(2) Kinetics parameters for As(III) adsorption onto ␥-Fe2 O3 –TiO2 nanoparticles in dark.
qe (mg/g) Kad (g/(mg h)) r2
(3) 1.31 7.06 0.9998
 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
Fig. 7. The pseudo-second-order rate kinetic model fitting of the adsorption kinetics of As(III) with ␥-Fe2 O3 –TiO2 nanoparticles in dark. Initial As(III) concentration 1.01 mg/L;
adsorbent dosage 0.5 g/L and solution pH 7.0 ± 0.1.
Fig. 8. Arsenic species concentrations vs treatment time under various conditions with initial As(III) concentration at 1.01 mg/L, adsorbent dosage 0.5 g/L and solution pH at
7.0 ± 0.1.
The changes of As(V) concentration under three different treat-
ments are shown in Fig. 8b. Under only UV light illumination, the
concentration of As(V) showed a slight increase with illumination
time up to 12 h due to the oxidation of As(III) when exposed to
the air. When ␥-Fe2 O3 –TiO2 nanoparticles were present without
UV light illumination, the concentration of As(V) decreased rapidly
due to its adsorption onto ␥-Fe2 O3 –TiO2 nanoparticles. After 12 h,
the concentration of As(V) decreased from 30.1 to 5.2 ␮g/L, rep-
resenting a 80.3% decrease. Under the presence of both UV light
and ␥-Fe2 O3 –TiO2 , the concentration of As(V) decreased initially,
which could be ascribed to the combination impacts from both
As(V) adsorption (decreasing the concentration of As(V)) and As(III)
oxidation (increasing the concentration of As(V)), while As(V)
adsorption plays the dominant role at the beginning of the inter-
action. As the illumination time increased, a slight increase of the
concentration of As(V) was observed. This increase could only be
ascribed to the continuous photooxidation of As(III) into As(V) by
the ␥-Fe2 O3 –TiO2 nanoparticles, which overrides the decrease of
the concentration of As(V) by adsorption. Thus, our experiment
provides clear evidence that in ␥-Fe2 O3 –TiO2 nanoparticles sus-
pension, As(III) can be oxidized into As(V) by dissolved O2 under UV
light, which thus eliminates the need for an expensive or possibly
hazardous oxidant for As(III) oxidation.
3.5. Effect of ionic strength and coexisting ions on As(III) removal
The effect of ionic strength on As(III) removal by ␥-Fe2 O3 –TiO2
is illustrated in Fig. 9. It was clear that when the ionic strength
increased from 0.001 to 0.1 M NaNO3 , the removal of As(III) almost
did not change. This is consistent with the fact that the surface
15
interaction between As(V) and iron is mainly inner sphere in
nature [35]. When changing the ionic strength of the system, non-
specifically adsorbed ions are more sensitive to the change of ionic
strength than specifically adsorbed ions because electrolytes can
also form outer-sphere complexes with the adsorbent through
electrostatic forces.
The arsenic removal by iron oxides was mainly realized by form-
ing surface complexes at the active sites on the surface of iron
oxides [34]. This is similar to that of ␥-Fe2 O3 –TiO2 nanoparticles.
So the removal of arsenic could be affected in the presence of those
ions which can compete with the anions of arsenic for the adsorp-
tive sites of adsorbent. Therefore, it is necessary to investigate the
competition between arsenic and commonly coexisting anions and
100
80
Arsenic removal (%)
60
40
20
0
1 2 3 4 5 6 7
0.001M NaNO3 0.01M NaNO3 0.1M NaNO3
Fig. 9. Effects of ionic strength on As(III) removal. Initial As(III) concentration
1.01 mg/L; adsorbent dosage 0.5 g/L; solution pH 7.0 ± 0.1 and reaction time 3 h.
16 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
Fig. 10. Effects of co-existing ions (1 mM) on As(III) removal at fixed initial As(III)
concentration 1.01 mg/L; adsorbent dosage 0.5 g/L; solution pH 7.0 ± 0.1 and reac-
tion time 3 h.
cations. As shown in Fig. 10 that the influence of Ca2+ and Mg2+
on the removal of As(III) were rather insignificant since they did
not compete with arsenic anions for the adsorptive sites of adsor-
bent at this pH [25]. SO4 2− , NO3 − and Cl− are chemically inert, and
the competing adsorption between these anions and arsenic can be
ignored. Compared with arsenic species, sulphate, nitrate and chlo-
ride are poor ligands and suggest a weaker adsorption mechanism
via outer-sphere complexation [25]. Phosphate, silicate, bicarbon-
ate could affect the removal of As(III), the effect of the anions
followed the selectivity pattern: phosphate > silicate > bicarbonate.
This result is in agreement with previous studies [34]. As located
in the same main group, the molecular structure and the chemical
property of arsenic and phosphate are very similarity. Phosphate
is the strongest competitor with arsenic anions for the adsorptive
sites on the surface of ␥-Fe2 O3 –TiO2 nanoparticles.
3.6. Regeneration studies
Considering the practical applicability, it is desirable that a
material should be fully recovered and reused so that it can be put
into cyclic use in a cost effective manner. Either alkali or strong
acids has been used to regenerate arsenic saturated adsorbents
including desorption of arsenic from ␥-Fe2 O3 –TiO2 nanoparticles.
To study the regeneration, ␥-Fe2 O3 –TiO2 nanoparticles were first
loaded with arsenic solution by stirring the suspension with initial
arsenite concentration of 1.01 mg/L and adsorbent dosage of 0.5 g/L
for 12 h under UV light. This process was repeated until the material
get saturated. Regeneration studies were proceeded by stirring the
required quantity of arsenic saturated ␥-Fe2 O3 –TiO2 with 0.1 M
NaOH solution for 3 h. In the process of adsorption, most As(III)
was oxidized to As(V) through photocatalytic oxidation. It has been
confirmed that As(V) removal was greatly affected by solution pH,
when the solution pH was 9.0 and the equilibrium concentration
of As(V) was below 0.7 mg/g, the As(V) adsorption capacity was
below 2.5 mg/g [24]. When the arsenic saturated ␥-Fe2 O3 –TiO2
nanoparticles were stirred with NaOH solution, As anions could
desorb from the surface of ␥-Fe2 O3 –TiO2 nanoparticles through
hydroxyl exchange, in addition, electrostatic repulsion may also
play an important role in the re-generation process. Fig. 11 shows
the removal efficiency of As(III) by ␥-Fe2 O3 –TiO2 nanoparticles
during five adsorption–regeneration cycles. It was observed that
in the first cycle, the removal efficiency decreased obviously, and
in the subsequent cycles, the removal efficiency decreased slowly.
However, the removal efficiency after five cycles can still reach
84%, showing that it could be used as an effective material for
many cycles by regeneration. It should be mentioned that, some
ultrapure water was used to rinse the surface of regenerated
adsorbent to remove unwanted ions after each cycle. When
95
Arsenic removal (%)
90
85
80
0 1 2 3 4 5
Cycle number
Fig. 11. Regeneration studies of ␥-Fe2 O3 –TiO2 nanoparticles after five cycles.
undergoing desorption process, the initial 1.01 mg/L of As anions
were concentrated into a smaller volume and thereby the final
concentration of As anions reached as high as 200 mg/L. It was also
found that ␥-Fe2 O3 –TiO2 nanoparticles agglomerated markedly
and the specific surface area decreased after each cycle. This can
explain the decrease of removal efficiency. After each cycle, the
adsorbent was crushed again to retain high specific surface area.
The results of regeneration studies suggest that ␥-Fe2 O3 –TiO2
nanoparticles can be used in a continuous flow for removal of
arsenite.
3.7. Mechanism of arsenite removal by -Fe2 O3 –TiO2
As the life-times of both photon-excited electron-hole pairs and
subsequently produced radicals such as • OH, O2 •− , and HO2 • are
short, the photocatalytic reactions occur only at/near the photocat-
alyst surface [3]. As can be seen from the section of “As(III) removal
through simultaneous photocatalytic oxidation and adsorption”,
the photocatalysts possess a good adsorption capability of As(III),
when the reactor was placed under UV light, most As(III) could
be oxidated to As(V). ␥-Fe2 O3 –TiO2 nanoparticles retained consid-
erable photocatalytic activity after 12 h photocatalytic oxidation,
this reflects that not all active sites of TiO2 nanoparticles have been
occupied by As(V) in the process of photocatalytic oxidation and
adsorption. It should be deduced that, with more active adsorption
sites, ␥-Fe2 O3 nanoparticles could compete for As(V) anions with
TiO2 nanoparticles, As(V) anions can desorb from the surface of TiO2
nanoparticles, thus most As(V) anions were adsorbed by ␥-Fe2 O3 .
From the above analysis, the mechanism of As(III) removal by
␥-Fe2 O3 –TiO2 is proposed as following three steps: (1) adsorption
of As(III) on the surface of ␥-Fe2 O3 –TiO2 ; (2) oxidation of As(III) to
As(V) and As(V) desorbed from the surface of TiO2 and (3) As(V)
adsorbed by ␥-Fe2 O3 –TiO2 . From the above discussion, the whole
process can be summarized by reaction Eqs. (4)–(6).
As(III)(aq) + (−SFe−Ti ) → As(III) − SFe−Ti (4)
hv
1
As(III) − SFe−Ti + O2 + 2 As(III)(aq)−→1 As(V)(aq) + 2 As(III) − SFe−Ti (5)
As(V)(aq) + (−SFe−Ti ) → As(V) − SFe−Ti (6)
where (−SFe–Ti ) represents an adsorption site on the ␥-Fe2 O3 –TiO2
nanoparticle surface. As(III)−SFe–Ti represents the As(III) surface
species and As(V)−SFe–Ti represents the As(V) surface species. This
photooxidation and sorption mechanism could quite well explain
the above experimental results.
4. Conclusions
Magnetic separation process provides a low-cost and environ-
mentally benign water treatment process. Consequently, a novel
 L. Yu et al. / Journal of Hazardous Materials 246–247 (2013) 10–17
Fe–Ti binary oxide nano-material with magnetic property was pre-
pared, characterized and applied for As(III) removal. The prepared
␥-Fe2 O3 –TiO2 nanoparticles possess of synergy of the photocat-
alytic ability of TiO2 (anatase) for oxidation of As(III) to As(V) and
the adsorption performance of ␥-Fe2 O3 for the removal of As(V).
The obtained magnetic nano-material can oxidize As(III) to As(V)
under UV light irradiation at various pH values. The arsenic uptake
was rapid initially up to 3 h and then gradually slowed down as
it was closed to equilibrium. The removal efficiency was not so
obviously affected by pH within limit, as acidic pH condition is
beneficial for As(V) adsorption whereas alkaline condition is favor-
able for photooxidation of As(III). The maximal removal capability
for As(III) was 33.03 mg/g at solution p H 7.0, initial As(III) con-
centration 50 mg/L. Among the ions tested, phosphate was the
greatest competitor with arsenic for adsorptive sites on the sur-
face of nano-material. Furthermore, ionic strength, the presence
of SO4 2− , NO3 − , Cl− , Ca2+ , and Mg2+ had no significant effect on
arsenic removal. The regeneration studies demonstrated that ␥-
Fe2 O3 –TiO2 nanoparticles can be recovered effectively, ensuring
the reusability of ␥-Fe2 O3 –TiO2 for arsenite removal. The well pho-
tocatalytic oxidation and adsorption properties of ␥-Fe2 O3 –TiO2
nanoparticles make them potentially attractive material for the
removal of arsenite from water.
Acknowledgments
This work was supported by the National Basic Research Pro-
gram of China (Grant No. 2011CB933700-G) and National Natural
Science Foundation of China (Grant No. 41273123).
References
[1] D.K. Nordstrom, Worldwide occurrences of arsenic in groundwater, Science
296 (2002) 2143–2144.
[2] M. Amini, K.C. Abbaspour, M. Berg, L. Winkel, S.J. Hug, E. Hoehn, H. Yang,
C.A. Johnson, Statistical modeling of global geogenic arsenic contamination in
groundwater, Environ. Sci. Technol. 42 (2008) 3669–3675.
[3] Q. Li, N.J. Eate, J.K. Shang, As(III) removal by palladium-modified nitrogen-
doped titanium oxide nanoparticle photocatalyst, Environ. Sci. Technol. 43
(2009) 1534–1539.
[4] M. Piril, M. Martikainen, K. Ainassaari, T. Kuokkanen, R.L. Keiski, Removal of
aqueous As(III) and As(V) by hydrous titanium dioxide, J. Colloids Interface. Sci.
353 (2011) 257–262.
[5] A.H. Smith, P.A. Lopipero, M.N. Bates, C.M. Steinmaus, Arsenic epidemiology
and drinking water standards, Science 296 (2002) 2145–2146.
[6] R.M. Dhoble, S. Lunge, A.G. Bhole, S. Rayalu, Magnetic binary oxide particles
(MBOP): a promising adsorbent for removal of As(III) in water, Water Res. 45
(2011) 4769–4781.
[7] US EPA, Technologies and costs for removal of arsenic from drinking water,
Washington, DC, 2000.
[8] M. Borho, P. Wilderer, Optimized removal of arsenite(III) by adaptation of oxi-
dation and precipitation processes to the filtration step, Water Sci. Technol. 34
(1996) 25–31.
[9] H. Lee, W. Choi, Photocatalytic oxidation of arsenite in TiO2 suspension: kinetics
and mechanisms, Environ. Sci. Technol. 36 (2002) 3872–3878.
[10] Y. Kim, C. Kim, I. Choi, S. Rengaraj, J. Yi, Arsenic removal using mesoporous
alumina prepared via a templating method, Environ. Sci. Technol. 38 (2004)
924–931.
[11] M. Kim, J. Nriagu, Oxidation of arsenite in groundwater using ozone and oxygen,
Sci. Total Environ. 247 (2000) 71–79.
17
[12] M. Pettine, L. Campanella, F.J. Millero, Arsenite oxidation by H2 O2 in aqueous
solutions, Geochim. Cosmochim. Acta 63 (1999) 2727–2735.
[13] M.C. Dodd, N.D. Vu, A. Ammann, V.C. Le, R. Kissner, H.V. Pham, T.H. Cao, M.
Berg, U.V. Gunten, Kinetics and mechanistic aspects of As(III) oxidation by aque-
ous chlorine, chloramines, and ozone: relevance to drinking water treatment,
Environ. Sci. Technol. 40 (2006) 3285–3292.
[14] V.Q. Chiu, J.G. Hering, Arsenic adsorption and oxidation at manganite surfaces.
1. Method for simultaneous determination of adsorbed and dissolved arsenic
species, Environ. Sci. Technol. 34 (2000) 2029–2034.
[15] H. Yang, W.Y. Lin, K. Rajeshwar, Homogeneous and heterogeneous photo-
catalytic reactions involving As(III) and As(V) species in aqueous media, J.
Photochem. Photobiol. A 123 (1999) 137–143.
[16] M.T. Emett, G.H. Khoe, Photochemical oxidation of arsenic by oxygen and iron
in acidic solutions, Water Res. 35 (2001) 649–656.
[17] S.J. Hug, L. Canonica, M. Wegelin, D. Gechter, U.V. Gunten, Solar oxidation and
removal of arsenic at circumneutral pH in iron containing waters, Environ. Sci.
Technol. 35 (2001) 2114–2121.
[18] B.D. Kocar, W.P. Inskeep, Photochemical oxidation of As(III) in ferrioxalate solu-
tions, Environ. Sci. Technol. 37 (2003) 1581–1588.
[19] S.J. Hug, O. Leupin, Iron-catalyzed oxidation of arsenic(III) by oxygen and by
hydrogen peroxide: pH-dependent formation of oxidants in the Fenton reac-
tion, Environ. Sci. Technol. 37 (2003) 2734–2742.
[20] J. Ryu, W. Choi, Effects of TiO2 surface modifications on photocatalytic oxi-
dation of arsenite: the role of superoxides, Environ. Sci. Technol. 38 (2004)
2928–2933.
[21] M.A. Ferguson, M.R. Hoffmann, J.G. Hering, TiO2 -photocatalyzed As(III) oxi-
dation in aqueous suspensions: reaction kinetics and effects of adsorption,
Environ. Sci. Technol. 39 (2005) 1880–1886.
[22] P.K. Dutta, S.O. Pehkonen, V.K. Sharma, A.K. Ray, Photocatalytic oxidation of
arsenic(III): evidence of hydroxyl radicals, Environ. Sci. Technol. 39 (2005)
1827–1834.
[23] W. Zhou, H.G. Fu, K. Pan, C.G. Tian, Y. Qu, P.P. Lu, C.C. Sun, Mesoporous
TiO2 /␣-Fe2 O3 , bifunctional composites for effective elimination of arsenite con-
tamination through simultaneous photocatalytic oxidation and adsorption, J.
Phys. Chem. CV 112 (2008) 19584–19589.
[24] T. Tuutijärvi, J. Lu, M. Sillanpää, G. Chen, As(V) adsorption on maghemite
nanoparticles, J. Hazard. Mater. 166 (2009) 1415–1420.
[25] J. Hu, G.H. Chen, M.C.L. Irene, Removal and recovery of Cr(VI) from wastewater
by maghemite nanoparticles, Water Res. 39 (2005) 4528–4536.
[26] D.N. Thanh, M. Singh, P. Ulbrich, N. Strnadova, F. Štěpánek, Perlite incorporat-
ing (-Fe2 O3 and ␣-MnO2 nanomaterials: preparation and evaluation of a new
adsorbent for As(V) removal, Sep. Purif. Technol. 82 (2011) 93–101.
[27] M.H. Zhou, J.G. Yu, B. Cheng, Effects of Fe-doping on the photocatalytic activ-
ity of mesoporous TiO2 powders prepared by an ultrasonic method, J. Hazard.
Mater. 137 (2006) 1838–1847.
[28] J.X. Li, J.H. Xu, W.L. Dai, H.X. Li, K.N. Fan, Direct hydro-alcohol thermal synthesis
of special core-shell structured Fe-doped titania microspheres with extended
visible light response and enhanced photoactivity, Appl. Catal. B: Environ. 85
(2009) 162–170.
[29] H.F. Liu, Z.G. Jia, S.F. Ji, Y.Y. Zheng, M. Li, H. Yang, Synthesis of TiO2 /SiO2 @Fe3 O4
magnetic microspheres and their properties of photocatalytic degradation
dyestuff, Catal. Today 175 (2011) 293–298.
[30] L. Cumbal, A.K. Sengupta, Arsenic removal using polymer-supported hydrated
iron(III) oxide nanoparticles: role of Donnan membrane effect, Environ. Sci.
Technol. 39 (2005) 6508–6515.
[31] Z.J. Li, S.B. Deng, G. Yu, J. Huang, V.C. Lim, As(V) and As(III) removal from water
by a Ce–Ti oxide adsorbent: behavior and mechanism, Chem. Eng. J. 161 (2010)
106–113.
[32] M.E. Pena, G.P. Korfiatis, M. Patel, L. Lippincott, X. Meng, Adsorption of
As(V) and As(III) by nanocrystalline titanium dioxide, Water Res. 39 (2005)
2327–2337.
[33] Y. Sag, Y. Aktay, Kinetic studies on sorption of Cr(VI) and Cu(II) ions
by chitin, chitosan and Rhizopus arrhizus, Biochem. Eng. J. 12 (2002)
143–153.
[34] G.S. Zhang, J.H. Qu, H.J. Liu, R.P. Liu, R.C. Wu, Preparation and evaluation of a
novel Fe–Mn binary oxide adsorbent for effective arsenite removal, Water Res.
41 (2007) 1921–1928.
[35] L. Zeng, A method for preparing silica-containing iron(III) oxide adsorbents for
arsenic removal, Water Res. 37 (2003) 4351–4358.