Ferromagnetic photocatalysts of FeTiO3-Fe2O3 nanocomposites

Article abstract of DOI:10.1039/c7ra11007c

Besides harvesting sunlight over a broad wavelength range as much as possible, the efficient separation of photo-generated electron-hole pairs is vital for the development of high-quality photocatalysts. In this work, we design FeTiO₃ (FTO) and Fe₂O₃ (FO) nanocomposites (xFTO-(1 - x)FO), which are simply prepared using a hydrothermal method. The FTO is epitaxially grown on FO nanoparticles, and with the increasing concentration of FTO, the band gaps decrease from 2.43 eV (x = 1.00) to 1.56 eV (x = 0.60). The photocatalytic capability is significantly improved such that xFTO-(1 - x)FO (x = 0.60) shows the highest value, which is about 8 times that of FO and 4 times that of FTO. Furthermore, strong ferromagnetism with saturated magnetization larger than 6 emu g^-1 is observed in xFTO-(1 - x)FO with x ≥ 0.60. xFTO-(1 - x)FO (x = 0.60) is further annealed at various temperatures. After annealing at 300 °C, the photocatalytic capability and ferromagnetism are both improved, by 38% and 31%, respectively, but drastically decrease with further increase of the annealing temperature to higher than 400 °C. The mechanism of the enhanced photocatalytic capability has been ascribed to the interdiffusion between FTO and FO at interfaces inside the nanocomposite particles, resulting in the formation of p-n junctions, which may facilitate the separation of photo-generated electron-hole pairs by the built-in-electric field. A significant enhancement of ferromagnetism occurs at the interdiffusion region with higher concentrations of FTO.

Full text of DOI:10.1039/c7ra11007c

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Ferromagnetic photocatalysts of FeTiO –Fe O
3
2 3
nanocomposites†
Cite this: RSC Adv., 2017, 7, 54594
a
a
a
a
a
a
Baizhi Gao,
Lujun Wei,
Caiping Yang, Jun Chen, Yuxing Ma, Jiachen Xie, Hao Zhang,
b
a
bc
ac
Qi Li,* Jun Du* and Qingyu Xu*
Besides harvesting sunlight over a broad wavelength range as much as possible, the efficient separation of
photo-generated electron–hole pairs is vital for the development of high-quality photocatalysts. In this
work, we design FeTiO
3
(FTO) and Fe
2
O
3
(FO) nanocomposites (xFTO–(1 ꢀ x)FO), which are simply
prepared using a hydrothermal method. The FTO is epitaxially grown on FO nanoparticles, and with the
increasing concentration of FTO, the band gaps decrease from 2.43 eV (x ¼ 1.00) to 1.56 eV (x ¼ 0.60). The
photocatalytic capability is significantly improved such that xFTO–(1 ꢀ x)FO (x ¼ 0.60) shows the highest
value, which is about 8 times that of FO and 4 times that of FTO. Furthermore, strong ferromagnetism with
ꢀ
1
saturated magnetization larger than 6 emu g is observed in xFTO–(1 ꢀ x)FO with x $ 0.60. xFTO–(1 ꢀ x)
ꢁ
FO (x ¼ 0.60) is further annealed at various temperatures. After annealing at 300 C, the photocatalytic
capability and ferromagnetism are both improved, by 38% and 31%, respectively, but drastically decrease
ꢁ
with further increase of the annealing temperature to higher than 400 C. The mechanism of the enhanced
Received 6th October 2017
Accepted 22nd November 2017
photocatalytic capability has been ascribed to the interdiffusion between FTO and FO at interfaces inside the
nanocomposite particles, resulting in the formation of p–n junctions, which may facilitate the separation of
DOI: 10.1039/c7ra11007c
photo-generated electron–hole pairs by the built-in-electric field.
A significant enhancement of
rsc.li/rsc-advances
ferromagnetism occurs at the interdiffusion region with higher concentrations of FTO.
Based on the above-mentioned factors for high-
performance photocatalyst, we develop novel nano-
(FTO) and a-Fe (FO),
Introduction
a
Photocatalysis is an environment-friendly energy transforming composite photocatalyst with FeTiO
O
2 3
3
process from solar energy to chemical energy, which has attrac- which has been proved to be a successful technique to develop
ted extensive research interest since the discovery of the rst efficient photocatalyst. The band gaps of FTO and FO are
1
17,18
photocatalyst (TiO
the band gap of TiO
2
) used for water splitting in 1972. However, ꢂ2.5 eV (ref. 16) and ꢂ2.2 eV,
respectively. And FTO might
is 3.2 eV for the anatase phase and 3.0 eV for have large adsorption and photocatalytic capability due to its
2
the rutile phase, which makes TiO nearly inactive under visible easier formation of vacancy defects on the surface, which
2
2
–4
light irradiation.
controlling the band gaps for better utilization of solar energy,
and improving carrier separation and transportation efficiency suitable to work under visible illuminations, their high
Many strategies were put forward for would both enhance the porous surface structure formation
5
19
2
and O diffusivity. Although the above factors make them
6,7
for enhancing chemical reactivity. The formation of hetero- carriers' recombination rate makes them not so efficient in
interfaces including p–n junctions and Schottky contacts has practice.
been proved to be very effective for carrier separation. For p–n eld in p–n junctions. It has been reported that opposite
junctions, the generation of a built-in electric eld at the inter- electrical properties can be achieved by changing the relative
8–12
4
17,18
This might be changed based on built-in-electric
13
faces by equaling the Fermi level leads to the transportation of concentration in FTO
x
–FO(1ꢀx) (in molar ratio) solid solutions,
20
electrons and holes in opposite directions and the recombination with p-type for x > 0.73 and n-type for x < 0.73. Thus, p–n
14,15
is therefore signicantly suppressed.
junction might be formed at the interface of FTO and FO
composites by proper interdiffusion. Furthermore, FTO and
2
1,22
FO can form solid solution in full concentration range,
and
a
School of Physics, Southeast University, Nanjing 211189, China. E-mail: qli@seu.edu.
cn; xuqingyu@seu.edu.cn
strong ferromagnetism has been observed when FTO concen-
tration is higher than 0.5, though pure FTO is antiferromag-
b
School of Physics, Nanjing University, Nanjing 210093, China. E-mail: jdu@nju.edu.
2
3
netic. Thus, this may further make xFTO–(1 ꢀ x)FO
cn
c
National Laboratory of Solid State Microstructures, Nanjing University, Nanjing nanocomposites to be potential ferromagnetic photocatalysts,
2
†
1
10093, China
which can be magnetically separable for the efficient separa-
tion and recovery from treated water.
Electronic supplementary information (ESI) available. See DOI:
0.1039/c7ra11007c
24–27
54594 | RSC Adv., 2017, 7, 54594–54602
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
In this paper, we simply prepared xFTO–(1 ꢀ x)FO nano- irradiation. Typically, 50 mg of photocatalyst was dispersed in
ꢀ1
composites by hydrothermal method. By carefully tuning the 150 ml of 2 mg l RhB aqueous solution. A 500 W Xe lamp
concentration of FTO, signicantly enhanced photocatalytic (Beijing Trusttech Co Ltd, CHF-XM) equipped with a visible
capability with strong ferromagnetism has been observed, pass lter (l ¼ 400–800 nm). The concentrations of RhB were
which has been conrmed to be due to the interdiffusion of FTO monitored at 554 nm by using a UV-vis DRS (Hitach U-3900). A
and FO at the interfaces inside the particles.
30 min adsorption in darkness was applied as the pretreatment
for absorption–desorption equilibrium before photocatalytic
process.
Experimental
xFTO–(1 ꢀ x)FO nanocomposites (x ¼ 0.00, 0.20, 0.40, 0.60, 0.80
Results and discussion
and 1.00 in molar ratio) were prepared by hydrothermal
21
method.
aqueous solution), titanium isopropoxide (TTIP, 95% aqueous
solution) and FeSO $7H O (99.9%) were used as the starting
Tetrabutylammonium hydroxide (TBAH, 40%
Fig. 1 shows the XRD patterns of xFTO–(1 ꢀ x)FO nano-
composites. FTO and FO are both in corundum structure (space
28
4
2
group R3 for FTO and R3c for FO), similar XRD patterns of pure
phases can be observed for all the samples as shown in Fig. 1(a).
A magnied view of (110) peaks is illustrated in the inset of
Fig. 1(b). FTO has larger lattice constants than FO, thus the
diffraction peaks gradually shi to smaller angles, as shown in
Fig. 1(b), indicating the lattice expansion with increasing
concentration of FTO calculated using Bragg's law. Interestingly,
ꢀ
ꢀ
materials for the preparation of FTO, and KOH (99.9%) was
used for the pH control. 0.005 mol TBAH (3.25 ml) was taken
and diluted in 40 ml deionized water, and then 0.005 mol TTIP
(
1.5312 ml) was added rapidly. The mixture was stirred by
ꢁ
a magnetic stirrer and kept at the temperature of 50 C, until it
4 2
was completely clear. 0.005 mol FeSO $7H O powder (1.39 g)
was dissolved in 10 ml deionized water and immediately poured
into the mixture which was kept on stirring. Then 0.05 mol KOH
(2.805 g) was added into the mixture directly to adjust the pH
value to 14. Aer that, the FO powders were added, and the
amount was adjusted according to the designed composition.
The stirring process was continued for another 20 min. The
mixture was then moved to an autoclave. The autoclave was
ꢁ
ꢁ
ꢀ1
gradually heated to 220 C in a speed of around 3–5 C min
ꢁ
and kept at 220 C for 10 h, then slowly cooled down to room
temperature. The precipitates were collected by centrifugation,
washed by deionized water and ethanol. The washed products
ꢁ
were nally dried at 70 C for 3 h. Annealing process was per-
formed at various temperatures for 3 hours. It must be noted
that air should be isolated by the protection of Ar atmosphere in
the whole preparation process.
The crystal structure was studied by q–2q powder X-ray
diffraction (XRD, Rigaku Smartlab3) using Cu Ka radiation
˚
(
1.5406 A). The morphologies were studied by a scanning elec-
tron microscope (SEM, FEI Inspect F50), equipped with an
energy dispersive X-ray spectroscope (EDX). Raman spectra were
carried out on a Horiba Jobin Yvon LabRAM HR 800 micro-
Raman spectrometer using a 532 nm laser. Band gaps were
determined from the UV-vis diffuse reectance spectrometer
(DRS, Hitach U-3900). X-ray photoelectron spectroscopy (XPS)
was carried out using an X-ray photoelectron spectrometer
(
ThermoFisher Scientic) with Al Ka X-ray source (hn ¼ 1486.6
eV). The samples for XPS measurements were kept in the high-
vacuum chamber overnight to remove the adsorbed air. Bru-
nauer–Emmett–Teller (BET) specic surface areas were deter-
mined from
N
2
adsorption by using an ASAP 2020
(Micrometrics USA) analyzer. The magnetic properties were
measured by a vibrating sample magnetometer (VSM, Micro-
sense EV7) with maximum applied eld of 10 kOe. All the
measurements were performed at room temperature.
The photocatalytic activities of the samples were tested by
the photocatalytic decomposition of model contaminates,
Fig. 1 (a) XRD patterns of xFTO–(1 ꢀ x)FO nanocomposites, the
standard XRD patterns of FTO and FO are shown on the bottom and
top, respectively. (b) The positions of (110) peaks in dependence on x.
The inset in (a) shows the images of the powders, and inset of (b)
Rhodamine B (RhB), at room temperature with visible light shows the magnified view of (110) peaks.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 54594–54602 | 54595

View Article Online
RSC Advances
Paper
a sudden shi of the diffraction peak can be observed between wrapped morphology can be clearly seen in xFTO–(1 ꢀ x)FO (x ¼
FTO and xFTO–(1 ꢀ x)FO (x ¼ 0.80), while the shi of the other 0.80) in Fig. 2(e), which is in agreement with its XRD pattern.
peaks is much smaller in a nearly linear way. The sudden shi of The exact concentration of FTO and FO in the nanocomposites
the diffraction peaks from pure FTO to xFTO–(1 ꢀ x)FO (x ¼ 0.80) is further measured by EDX, and the results are listed in Table 1.
might be due to the epitaxial growth of FTO on FO particle It can be clearly seen that the atomic ratio between Fe and Ti is
surface, since FTO and FO have the similar crystalline structure. very close to 1 for pure FTO. However, slight deviation between
This can be evidenced by the observation of the superimposed the measured concentration and designed concentration of
peaks of (110) in xFTO–(1 ꢀ x)FO (x ¼ 0.80) from FO and FTO. In FTO in the nanocomposites can be observed, which might be
xFTO–(1 ꢀ x)FO nanocomposites, FTO is grown on the surfaces due to the slight uctuation in ratios of starting materials and
of FO particles and the lattice of FTO is constrained by that of FO. incomplete fabrication of FTO during hydrothermal process.
With increasing x, the thickness of FTO layer increases, thus the For simplicity, the designed concentration of FTO is used in the
lattice gradually relaxes, leading to the shi of diffraction peaks paper to denote each sample.
to smaller angle. For xFTO–(1 ꢀ x)FO (x ¼ 0.80), some FTO layers
The structures of xFTO–(1 ꢀ x)FO nanocomposites were
on FO particles are thick enough, and the lattice eventually relax further studied by Raman Spectra and UV-vis DRS, and the
to that of FTO. The crystallite size of xFTO–(1 ꢀ x)FO calculated results are shown in Fig. 3. It has been reported that the Raman
ꢀ
1
using Scherrer's equation is listed in Table S1 (ESI†), which is in spectra of FO have two A1g modes (225 and 498 cm ) and ve E
g
ꢀ
1
30
the range between 40 nm and 60 nm. The images of xFTO–(1 ꢀ x) modes (247, 293, 299, 412, and 613 cm ). The ve Raman
FO powders are shown in inset of Fig. 1(a). As can be seen, the modes corresponding to FO can be clearly resolved, which
ꢀ
1
pure FTO powders shows gray black color, and FO shows red locate at 216, 280, 400, 490, and 600 cm . With increasing x,
brown color. A gradual transformation between these two colors the peak position shows little variation, indicating that the
with variation of the relative concentration of FTO and FO can be preparation process has little inuence on the structure of FO
clearly observed. This clearly conrms the coverage of FTO on FO. particles. It can also be seen that no extra Raman modes can be
Fig. 2 shows the morphologies of xFTO–(1 ꢀ x)FO nano- observed with increasing x. This might be due to that the
composites observed by SEM. As can be seen, the morphology of Raman signal from FTO is quite weak, leading to the coverage of
pure FTO particles are mainly nanodiscs (Fig. 2(f)), which is due small peaks by noises. Furthermore, most of the Raman modes
With only of FTO are close to those of FO, and the peaks are quite broad,
slight introduction of FO with concentration of 0.20, the which makes them difficult to be separated. However, a peak
morphology changes drastically to particle, which is quite locating at 673 cm corresponding to the A Raman mode of
similar with the morphology of FO particles. This can be FTO can be clearly resolved in the Raman spectrum of FTO. In
21,29
to the predominantly exposed {0001} polar facets.
31
ꢀ1
g
29
understood by the mechanism of epitaxial growth of FTO on FO contrast to the nearly unshied Raman modes from FO, a clear
ꢀ
1
particles, which wraps the surfaces of FO particles. And a fully shi to lower wavenumber (664 cm ) can be observed in xFTO–
Fig. 2 (a), (b), (c), (d), (e) and (f) are SEM images of xFTO–(1 ꢀ x)FO with x ¼ 0.00, 0.20, 0.40, 0.60, 0.80, 1.00, respectively.
54596 | RSC Adv., 2017, 7, 54594–54602
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Table 1 The measured concentration of xFTO–(1 ꢀ x)FO nanocomposites by EDX
Samples (xFTO–(1 ꢀ x)FO)
x ¼ 1.00
x ¼ 0.80
x ¼ 0.60
x ¼ 0.40
x ¼ 0.20
Fe K (%)
Ti K (%)
Measured ratio (Fe/Ti)
Measured x
50.74
49.26
1.03
—
55.16
44.84
1.23
72.72
27.28
2.67
81.82
18.18
4.50
89.10
10.90
8.17
0.90
0.55
0.36
0.22
n
33,34
the Kubelka–Munk equation: (ahn) ¼ A(hn ꢀ E ),
where A is
g
a constant, hn is the photon energy, E is the band gap of the
g
semiconductor, a is the absorption coefficient, and n is 0.5 for
32
indirect band gap materials, such as FTO. It has been reported
35
that there are direct and indirect band gaps for FO. We tted
the absorbance curves of nanocomposites of x $ 0.60 using
Kubelka–Munk equation with n ¼ 0.5 and 2, and found the best

t can be obtained by using n ¼ 0.5 (indirect band gap for FTO),
as shown in the inset of Fig. 3(b). This is understandable that
FTO takes the main absorption of light in the nanocomposites
with high FTO concentration due to the main coverage of FTO
on FO particle surfaces. And this can also be explained as FTO
particles have lower reectance than FO, which plays a more
21
important role in the absorption of light in nanocomposites.
The band gap for FTO is ꢂ2.43 eV, which is consistent with the
17,20
previous reported value of ꢂ2.5 eV.
With decreasing x in the
nanocomposites, the band gaps decrease to 1.88 eV (x ¼ 0.80)
and 1.56 eV (x ¼ 0.60), leading to an enhanced absorption
17
capability of light through a broader wavelength range.
The valence states of Fe and Ti of nanocomposites (x ¼ 0.00,
0
.60, 0.80 and 1.00) was studied by XPS as in Fig. 4. The C 1s
peak (284.8 eV) is used for charge correction. It is reported that
2
+
3+
3+
4+
the binding energy of Fe , Fe , Ti and Ti are 709.0 eV,
36,37
7
11.0 eV, 457.1 eV and 458.6 eV, respectively,
which have
been used for peak ttings. The Fe 2p spectra are shown in
Fig. 4(a)–(d) from x ¼ 0.00, 0.60, 0.80 and 1.00, respectively. As
can be seen, with the introduction of FTO, the 2p3/2 peaks shi
3
+
to lower binding energy from 711.0 eV for Fe to 709.0 eV for
2+
32
Fe . From the peak tting of 2p3/2 tting, an obviously varied
proportion is observed, which is in agreement with the
measured ratios of xFTO–(1 ꢀ x)FO nanocomposites. The
results are shown in Table S2 (ESI†). The Ti 2p spectra are
shown in Fig. 4(e)–(g) for x ¼ 0.60, 0.80 and 1.00, respectively.
Fig. 3 (a) Raman spectra, and (b) UV-vis DRS of xFTO–(1 ꢀ x)FO
nanocomposites. The inset in (b) shows the curve fitting by Kubelka–
Munk function.
3
+
4+
There is only slight uctuation of the proportion of Ti and Ti
due to the different samples with slightly uctuated fabricating
conditions of FTO. Since XPS is a very sensitive surface tech-
nique with depth limitation of about 5 nm, the concentration of
Fe will increase drastically and be much higher than x due to
the coverage of FTO on FO particles without interdiffusion.
Thus the XPS results clearly conrm the interdiffusion between
FTO and FO.
(
1 ꢀ x)FO (x ¼ 0.80) for this mode of FTO. It has been reported
that this mode of FTO corresponds to Ti–O stretching motion,
which can qualitatively reect the distortion of TiO . This can
6
2
+
31
be understood by the growth of FTO on FO particle surface,
which induces lattice strain in FTO due to the epitaxial growth.
The band gaps were determined by UV-vis DRS patterns, as
shown in Fig. 3(b). It can be clearly seen that nanocomposites of
x # 0.40 show nearly constant absorbance in the full visible
light range, which means that the band gap cannot be deter-
mined in this method.
0
absorption near the band edge of a semiconductor oen obeys
Photocatalytic efficiencies were measured by decomposing
RhB as model contaminant under a visible light illumination at
38,39
room temperature.
The typical time-dependent UV-vis spectra
of RhB solution decomposed by xFTO–(1 ꢀ x)FO (x ¼ 0.60) in
photochemical reaction is shown in Fig. 5(a). The intensity of the
characteristic absorption peak of RhB solution decreases with
17,22
Only for the nanocomposites of x $
.60, a clear absorption edge can be resolved. The optical
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 54594–54602 | 54597

View Article Online
RSC Advances
Paper
Fig. 4 The XPS spectra of the nanocomposites forth Fe 2p of (a), (b), (c) and (d) for x ¼ 0.00, 0.60, 0.80, 1.00, respectively, and Ti 2p spectra of (e),
f) and (g) for x ¼ 0.60, 0.80, 1.00, respectively.
(
54598 | RSC Adv., 2017, 7, 54594–54602
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
time. The absorption peak position shis to shorter wavelength, photocatalytic capability is only similar to pure FTO and FO
revealing the mechanism of N-deethylation mechanism in the (Fig. S2, ESI†), indicating the important role of the interface
40
photochemical reaction. Photocatalytic processes for the region in the xFTO–(1 ꢀ x)FO (x ¼ 0.60) nanocomposites. Due the
degradation of RhB aqueous solutions using xFTO–(1 ꢀ x)FO are inorganic nature of our composites and photocatalytic operation
shown in Fig. 5(b). It can be clearly seen that xFTO–(1 ꢀ x)FO (x ¼ at room temperature, the good recycling ability can be
4
1,42
0
9
.60) shows the highest photocatalytic capability. Aer 5 h, about expected.
0% RhB was degraded. For better comprehension of kinetics of
The eld dependent magnetization (M ꢀ H) curves for xFTO–
the RhB degradation, the tting using pseudo-rst-order model (1 ꢀ x)FO were measured by VSM, as shown in Fig. 6. Clear
(ln(C/C
0
)) ¼ Kt, where t is the irradiation time and K is the hysteresis loops can be observed, indicating the ferromagnetic
reaction coefficient, is drawn in Fig. 5(c). The reaction coefficient nature. The saturated magnetization (Msat) in dependence on x
K calculated from the slope of the curves is illustrated in Fig. 5(d). is plotted in the inset of Fig. 6(a). With x # 0.40, only weak
The highest value of K of 0.375 is obtained for xFTO–(1 ꢀ x)FO (x ferromagnetism with small Msat is observed. While x is larger
¼
0.60), which is almost 8 times of FO (0.048), and 4 time of FTO than 0.60, strong ferromagnetism with large Msat can be seen.
23
(
0.086). The decrease of K for x ¼ 0.80 sample can be explained as Similar phenomenon has been reported by Y. Ishikawa et al.
excessive coverage of FO particles by FTO, which might impede Interestingly, considerably strong ferromagnetism is also
photo-generated electrons in FO to participate the reaction. observed for pure FTO, which is in contrasts to its antiferro-
23
Furthermore, the conductivity of xFTO–(1 ꢀ x)FO solid solutions magnetic nature. This might be due to the O deciency
has been studied, and the resistivity is minimum when x is observed by XPS, which can signicantly enhance the ferro-
between 0.50 and 0.60. This indicates the highest electric magnetism in FTO. The strongly enhanced ferromagnetism in
20
23
conductivity for xFTO–(1 ꢀ x)FO (x ¼ 0.60), which is benecial for solid solution of FTO and FO with FTO concentration higher
2
+
the photocatalytic activity. The specic surface area of xFTO–(1 ꢀ than 50% mainly comes from Fe ions and the magnetic
2
ꢀ1
3+
43
x)FO (x ¼ 0.60) (14.966 m
g ) is smaller than that of FTO moments of Fe ions are canceled each other out. However,
2
ꢀ1
(
24.554 m g ), excluding the contribution of surface area on the the saturated magnetization of xFTO–(1 ꢀ x)FO is much smaller
improvement of photocatalytic capability (Fig. S1, ESI†). We than that of solid solution with the same concentration. For
further prepared the mechanically mixed powder of FTO and FO example, Msat for solid solution with x ¼ 0.60 is larger than 150
ꢀ
3
ꢀ1
with molar ratio same to xFTO–(1 ꢀ x)FO (x ¼ 0.60), the emu cm , which is about 30 emu g (considering the average
Fig. 5 (a) Absorption spectra for the degradation of RhB using xFTO–(1 ꢀ x)FO (x ¼ 0.60). (b) The photocatalytic performances of xFTO–(1 ꢀ x)
FO under visible irradiation. (c) The pseudo-first-order kinetics fitting. (d) The reaction efficiency in dependence on x.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 54594–54602 | 54599

View Article Online
RSC Advances
Paper
Fig. 6 (a) M ꢀ H curves, and (b) the enlarged view of M ꢀ H curves for xFTO–(1 ꢀ x)FO. The inset in (a) shows the dependence of saturated
magnetization on x.
ꢀ
3
mass density of about 5 g cm ), while Msat of xFTO–(1 ꢀ x)FO (x which is 6 times of FTO and 11 times of FO, with strong room
ꢀ
1
¼
0.60) is only 6.3 emu g . Thus, we can conclude that in the temperature ferromagnetism, which makes it a potential
as-prepared xFTO–(1 ꢀ x)FO, interdiffusion happens and solid ferromagnetic photocatalyst. To understand the mechanism,
solution forms at interface region between FTO and FO with a schematic diagram is shown in Fig. 8. Due to the interdiffu-
varying FTO concentration from high at FTO side to low at FO sion at the interface between FTO and FO inside the nano-
side. The wasp-waist shape M ꢀ H curves shown in the composite particles, solution with high concentration of FTO
magnied view of low eld in Fig. 6(b) (x ¼ 0.80, red) is due to
the superposition of M ꢀ H curves of the solid solution with
44
varying FTO concentration, which has various coercivities.
To further improve the interdiffusion at the interface
between FTO and FO, annealing was performed to xFTO–(1 ꢀ x)
FO (x ¼ 0.60), which shows the best photocatalytic capability.
The XRD patterns show that no observable changes can be
ꢁ
observed with annealing temperature increased up to 700 C
(Fig. S3†). The SEM images show that the roughly wrapped
surface with FTO nanoparticles on FO particles disappears and
becomes smooth, which is similar to FO (Fig. S4†). This indi-
cates that the increasing annealing temperature improves the
interdiffusion between FTO and FO. Raman spectra were also
studied that no obvious difference appears among the annealed
samples, in agreement with XRD (Fig. S5†). The valence states of
ꢁ
Fe 2p and Ti 2p of annealed sample of x ¼ 0.60 at 300 C and
ꢁ
7
00 C were tested. No obvious change of valence states can be
ꢁ
observed at 300 C. But it shows a signicant transformation of
2
+
4+
3+
3+
ꢁ
21
Fe + Ti to Fe + Ti at 700 C (Fig. S6†). This might be
induced by many reasons including interdiffusion, valence
states transfer in FTO and O defects, which is not involved in
our discussion. The photocatalytic capability was measured for
the annealed samples by degradation of RhB (Fig. S7†). With
ꢁ
increasing the annealing temperature to 300 C, the reaction
coefficient K is further increased to 0.52, and abruptly drops
ꢁ
with further increasing annealing temperature above 400 C, as
shown in Fig. 7(a). Interestingly, the magnetic properties show
the similar dependence on annealing temperature, as shown in
ꢁ
Fig. 7(b). With increasing annealing temperature to 300 C, Msat
ꢀ1
increases to 8.1 emu g , which is about 31% increment. With
further increasing annealing temperature, Msat then decreases
(Fig. S8†).
This signicant enhanced photocatalytic capability has been
observed for xFTO–(1 ꢀ x)FO (x ¼ 0.60) annealed at 300 C,
Fig. 7 (a) The reaction coefficient K, and (b) saturated magnetization
ꢁ
Msat of xFTO–(1 ꢀ x)FO (x ¼ 0.60) annealed at various temperatures.
54600 | RSC Adv., 2017, 7, 54594–54602
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
increasing the annealing temperature. The mechanism has
been ascribed to the interdiffusion between FTO and FO at
interfaces inside the nanocomposite particles. The solid solu-
tion with various FTO concentrations at interface region is
formed by interdiffusion, leading to the formation of p–n
junction, which may facilitate the separation of photon-induced
electron–hole pairs by the built-in-electric eld. The strong
ferromagnetism comes from the solid solution at interdiffusion
regions with higher concentration of FTO.
Conflicts of interest
There are no conicts to declare.
Fig. 8 The schematic diagram of the mechanism of enhanced pho-
tocatalytic capability for the nanocomposites.
Acknowledgements
This work is supported by the National Natural Science Foun-
dation of China (51771053, 51471085), the Natural Science
Foundation of Jiangsu Province of China (BK20151400,
BK20140054), and the open research fund of Key Laboratory of
MEMS of Ministry of Education, Southeast University.
and low concentration of FTO forms at the interface. It has been
reported that electric conductivity of FTO and FO solid solution
shows p-type for FTO concentration higher than 73% and n-type
20
for FTO concentration lower than 73%. Thus, p–n junction
may form at the interface region, which induces the built-in-
electric eld at the interface. The electric eld may facilitate
the separation of the photon-generated electron–hole pairs and
References
15
avoid the recombination, which improves the photocatalytic
capability. And due to the existence of FTO–FO solid solution at
the interface with high FTO concentration, signicantly
1 A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
2 A. Manikandan, E. Manikandan, B. Meenatchi, S. Vadivel,
S. K. Jaganathan, R. Ladchumananandasivam, M. Henini,
M. Maaza and J. S. Aanand, J. Alloys Compd., 2017, 723,
1155–1161.
23
enhanced ferromagnetism can be observed. The proper
annealing may enlarge the interdiffusion region of FTO and FO
nanocomposites, which provides a larger separating region.
Thus, larger amounts of photon-generated carriers may partic-
ipate in the reaction without recombination. And the enlarged
interdiffusion region with high FTO concentration may further
increase the ferromagnetic properties. However, with further
increasing annealing temperature, the interdiffusion may
induce the uniform distribution of FTO in FO, which may
destroy the p–n junction. Thus, the separation of electrons and
holes is weakened. Furthermore, the solid solution with high
concentration of FTO decreases, leading to the decreases of
3 N. C. S. Selvam, A. Manikandan, L. J. Kennedy and J. J. Vijaya,
J. Colloid Interface Sci., 2013, 389, 91–98.
4 D.
Maruthamani,
S.
Vadivel,
M.
Kumaravel,
B. Saravanakumar, B. Paul, S. S. Dhar, A. Habibi-Yangjeh,
A. Manikandan and G. Ramadoss, J. Colloid Interface Sci.,
2017, 498, 449–459.
5 Z. S. Wang, Y. Cui, K. Hara, Y. Dan, C. Kasada and A. Shinpo,
Adv. Mater., 2007, 19, 1138–1141.
6 Z. Zhao, H. Liu and S. Chen, Nanoscale, 2012, 4, 7301–7373.
7 Q. D. Truong, J. Y. Liu, C. C. Chung and Y. C. Ling, Catal.
Commun., 2012, 19, 85–89.
23
magnetization.
8
X. Meng, Z. Li, H. Zeng, J. Chen and Z. Zhang, Appl. Catal., B,
017, 210, 160–172.
2
Conclusions
9
Y. Yuana, Z. Yub, Y. Li, H. Lu, X. Chen, W. Tu, Z. Ji and
In conclusion, xFTO–(1 ꢀ x)FO nanocomposites were prepared
Z. Zou, Appl. Catal., B, 2016, 181, 16–23.
using hydrothermal method. The structural characterization 10 H. Li, Y. Sang, S. Chang, X. Huang, Y. Zhang, R. Yang,
indicates the epitaxial growth of FTO on the surface of FO
H. Jiang, H. Liu and Z. Wang, Nano Lett., 2015, 15, 2372–
particles. The band gap decreases from 2.43 eV (x ¼ 1.00) to
2379.
1
.56 eV (x ¼ 0.60), which broadens the absorption wavelength 11 X. Meng and Z. Zhang, J. Photochem. Photobiol., A, 2015, 310,
range. The photocatalytic capability has been signicantly
33–44.
improved. xFTO–(1 ꢀ x)FO (x ¼ 0.60) shows the highest reaction 12 X. Yu, J. Zhang, Z. Zhao, W. Guo, J. Qiu, X. Mou, A. Li,
coefficient, which is about 8 times of FO and 4 times of FTO. J. P. Claverie and H. Liu, Nano Energy, 2015, 16, 207–217.
Furthermore, strong ferromagnetism has been observed for 13 H. Wang, L. Zhang, Z. Chen, J. Hu, S. Li, Z. Wang, J. Liu and
ꢀ
1
xFTO–(1 ꢀ x)FO with x $ 0.60, up to 7.27 emu g (x ¼ 0.80).
X. Wang, Chem. Soc. Rev., 2014, 43, 5234–5244.
The photocatalytic capability and ferromagnetism of xFTO–(1 ꢀ 14 L. Li, P. A. Salvador and G. S. Rohrer, Nanoscale, 2014, 6, 24.
x)FO (x ¼ 0.60) are both further improved by over 30% aer 15 H. Wang, L. Zhang, Z. Chen, J. Hu, S. Li, Z. Wang, J. Liu and
ꢁ
annealing at 300 C, but drastically decrease with further
X. Wang, Chem. Soc. Rev., 2014, 43, 5234.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 54594–54602 | 54601

View Article Online
RSC Advances
Paper
16 G. Rollmann, A. Rohrbach, P. Entel and J. Hafne, Phys. Rev. 32 D. Gu, Y. Qin, Y. Wen, L. Qin and H. J. Seo, J. Taiwan Inst.
B: Condens. Matter Mater. Phys., 2004, 69, 165107.
Chem. Eng., 2017, 78, 431.
17 D. S. Ginley and M. A. Butler, J. Appl. Phys., 1977, 48, 5.
18 M. Mishra and D. Chun, Appl. Catal., A, 2015, 498, 126–141.
19 L. Qin, Z. Cheng, J. A. Fan, D. Kopechek, D. Xu,
33 B. Meenatchi, V. S. Lakshmi, A. Manikandan, V. Renuga,
A. Sharmila, K. R. N. Deve and S. K. Jaganathan, J. Inorg.
Organomet. Polym., 2017, 27, 446–454.
N. Deshpande and L. Fan, J. Mater. Chem. A, 2015, 3, 11302. 34 S. Suguna, S. Shankar, S. K. Jaganathan and A. Manikandan,
0 Y. Ishikawa, J. Phys. Soc. Jpn., 1958, 13, 1. J. Supercond. Novel Magn., 2017, 30, 691–699.
1 Y. J. Kim, B. Gao, S. Y. Han, M. H. Jung, A. K. Chakraborty, 35 C. S. Garoufalis, P. Poulopoulos, N. Bouropoulos,
2
2
T. Ko, C. Lee and W. I. Lee, J. Phys. Chem. C, 2009, 113,
9179–19184.
2 X. Zhang, T. Li, Z. Gong, H. Zhao, L. Wang, J. Wan, D. Wang,
X. Li and W. Fu, J. Alloys Compd., 2015, 653, 619–623.
3 Y. Ishikawa and S. Akimoto, J. Phys. Soc. Jpn., 1957, 12, 10.
A. Barnasas and S. Baskoutas, Phys. E, 2017, 89, 67.
36 T. Yamashita and P. Hayes, Appl. Surf. Sci., 2008, 254, 2441–
2449.
37 M. C. Biesinger, L. W. M. Lau, A. R. Gerson and
R. S. C. Smart, Appl. Surf. Sci., 2010, 257, 887–898.
1
2
2
2
4 C. Barathiraja, A. Manikandan, A. M. U. Mohideen, 38 A. Manikandan, E. Hema, M. Durka, M. A. Selvi, T. Alagesan
S. Jayasree and S. A. Antony, J. Supercond. Novel Magn.,
016, 29, 477–486.
and S. A. Antony, J. Inorg. Organomet. Polym., 2015, 25, 804–
815.
2
25 V. M. Teresita, A. Manikandan, B. A. Josephine, S. Sujatha 39 G. Mathubala, A. Manikandan, S. A. Antony and P. Ramar, J.
and S. A. Antony, J. Supercond. Novel Magn., 2016, 29,
691–1701.
6 A. Manikandan, M. Durka, K. Seevakan and S. A. Antony, J.
Mol. Struct., 2016, 1113, 79–87.
40 Q. Wang, C. Chen, D. Zhao, W. Ma and J. Zhao, Langmuir,
2008, 24, 7338–7345.
1
2
2
2
2
3
3
Supercond. Novel Magn., 2015, 28, 1405–1416.
7 D. K. Manimegalai, A. Manikandan, S. Moortheswaran and
S. A. Antony, J. Supercond. Novel Magn., 2015, 28, 2755–2766. 42 S. Sivakumar, A. Selvaraj, A. K. Ramasamy and
41 Y. Guo, G. Zhang, J. Liu and Y. Zhang, RSC Adv., 2013, 3,
2963.
8 R. J. Harrison and S. A. T. Redfern, Phys. Chem. Miner., 2001,
8, 399–412.
9 X. Guang, J. Zheng, M. Zhao, L. Li and G. Li, RSC Adv., 2013,
, 13635.
V. Balasubramanian, Water, Air, Soil Pollut., 2013, 224, 1529.
43 H. Hojo, K. Fujita, H. Ikeno, T. Matoba, T. Mizoguchi,
I. Tanaka, T. Nakamura, Y. Takeda, T. Okane and
K. Tanaka, Appl. Phys. Lett., 2014, 104, 112408.
2
3
0 D. L. A. Faria, S. V. Silva and M. T. Oliveira, J. Raman 44 A. A. Crist ´o bal, C. P. Ramos, P. G. Bercoff, P. M. Botta and
Spectrosc., 1997, 28, 873.
1 X. Wu, S. Qin and L. Dubrovinsky, J. Solid State Chem., 2010,
J. M. P. L ´o pez, Mater. Chem. Phys., 2012, 133, 971.
183, 2483.
54602 | RSC Adv., 2017, 7, 54594–54602
This journal is © The Royal Society of Chemistry 2017