Synthesis of carbon materials-TiO2 hybrid nanostructures and their visible-light photo-catalytic activity

Article abstract of DOI:10.1002/cplu.201300380

Activated carbon, graphene, carbon nanotubes and fullerene were incorporated into TiO₂ by a solvothermal approach and thermal annealing to produce carbon materials-TiO₂ hybrid nanostructures. The carbon materials-TiO₂ prod

Full text of DOI:10.1002/cplu.201300380

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300380
Synthesis of Carbon Materials–TiO Hybrid Nanostructures
2
and Their Visible-Light Photo-catalytic Activity
[
a]
[b]
[c]
[b]
[d]
Qian Li, Juncao Bian, Li Zhang, Ruiqin Zhang, Guozhong Wang, and
[a]
Dickon H. L. Ng*
Activated carbon, graphene, carbon nanotubes and fullerene
these nanocomposites over that for only using pure TiO . The
2
were incorporated into TiO by a solvothermal approach and
superiority in photo-catalytic activities on the RhB molecules
resulted from contributions from the excellent adsorption
property, favourable chemical bond formation (TiꢀOꢀC), nar-
rower band gap, smaller particle size and effective charge-carri-
er separation of the nanocomposites. Compared with the gra-
2
thermal annealing to produce carbon materials–TiO2 hybrid
nanostructures. The carbon materials–TiO products were char-
2
acterised by using SEM, TEM, XRD, Raman spectroscopy, X-ray
photo-electron spectroscopy, UV-visible spectroscopy and
photo-luminescence. The aim was to study the interactions be-
phene–, carbon nanotube– and fullerene–TiO products, acti-
2
tween the main TiO phase and the carbon components, and
vated carbon–TiO2 exhibited weaker interactions between
carbon and titania, lower adsorption for RhB and a larger band
gap, which led to lower photo-catalytic activity of RhB.
2
the relationships between morphology, structure and photo-
degradation of the rhodamine B (RhB) dye. An enhanced
photo-catalytic degradation of RhB was achieved when using
Introduction
[
5]
[6]
[7]
Energy crisis and environmental deterioration are by-products
of industrialisation. As a remedy, photo-catalysis has been
doping, metal doping, coupled semiconductor and noble-
[2a,8]
metal-based composites
have been investigated. In addi-
[1]
widely employed to deal with air and water treatment. It is
tion, mesoporous hollow TiO structures with high crystallinity
2
[2b,c]
[1b]
[9]
known that anatase titania (TiO ) is a promising photo-catalytic
and surface areas,
TiO nanorod, and TiO nanotubes,
2 2
2
candidate material owing to its photo-stability, non-toxicity, rel-
were prepared in an effort to enhance mass transfer between
the active sites and to favour isolation between active sites of
the catalyst and the dye molecules.
[
2]
atively high catalytic performance and low cost. However,
anatase TiO suffers from rapid electron–hole recombination,
2
[
3]
which leads to low photo-catalytic efficiency. Moreover, TiO2
has a wide band gap (3.2 eV) and is only active under UV-light
irradiation, although UV radiation only makes up 5% of normal
Recently, photo-catalysts containing carbon components
have attracted a great deal of attention owing to their superior
adsorption ability for pollutants in addition to their photo-cata-
[
4]
[4,10]
sunlight. To use sunlight to activate TiO , measures have to
lytic activities.
For example, an activated carbon (AC) com-
2
be taken to broaden visible-light harvesting and retard the re-
combination process of electron–hole pairs for anatase after
their generation. Some approaches such as non-metal
ponent with a porous structure can provide a large specific
surface area for adsorption and, at the same time, serves as
[11]
a support to the catalyst. This type of product will increase
catalyst adsorption performance to pollutants as well as facili-
tate mass transfer during the photo-catalytic reaction. Another
example is that owing to the unique electric and structural
properties of carbon nanotubes (CNTs), which have high con-
ductivity, large specific surface area and strong adsorption ca-
pability. CNTs were considered to be suitable dopants for the
[a] Q. Li, Prof. D. H. L. Ng
Department of Physics
The Chinese University of Hong Kong
Shatin (Hong Kong)
Fax: (+852)39436392
E-mail: dng@phy.cuhk.edu.hk
[4]
[b] J. Bian, Prof. R. Zhang
catalyst to enhance photo-catalysis. One interesting report
was that CNTs had a large electron-storage capacity (one elec-
tron per 32 carbon atoms), and thus, could serve as excellent
Department of Physics and Materials Science
and Centre for Functional Photonics (CFP)
City University of Hong Kong (Hong Kong)
[12]
electron acceptors. Xu et al. synthesised CNT/TiO nanocom-
[c] Prof. L. Zhang
2
Department of Mechanical and Automation Engineering
The Chinese University of Hong Kong (Hong Kong)
posites, which were utilised to degrade gas-phase benzene, by
[13]
using a simple impregnation method. The functions of CNTs
were multi-fold, for example, one CNT could act as electron
traps to prolong the lifetime of the created electron–hole pairs
and the other acted a dispersing agent to control the mor-
phology of the nanocomposite. The formation of TiꢀC and Tiꢀ
[
d] Prof. G. Wang
Institute of Solid State Physics
Chinese Academy of Sciences
Anhui (P. R. China)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201300380.
OꢀC bonds on the surface of CNT–TiO film would narrow the
2
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 454

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[
4]
band gap of the nanocomposite. In another study it was
shown that, for fullerene (C ), with a high electron affinity and
60
a Fermi level located below the conduction band (CB) of TiO2,
its photo-generated electron would be energetically favoured
[
3]
to flow from the CB of TiO to C . It is also known that gra-
2
60
phene (GR) and reduced graphene oxide (GO) are outstanding
[
1a,10a,14]
species among the carbon allotropes.
GR possesses
a high mobility of charge carriers at room temperature
2
ꢀ1 ꢀ1
(
200000 cm V S ) and exhibits an extremely high theoretical
2
ꢀ1 [15]
specific surface area (ꢁ2600 m g ); thus GR-based photo-
catalysts have been extensively applied to the photo-catalytic
[
14d,16]
degradation of organic compounds.
GO is often used as
a precursor of GR in the preparation of GR-based photo-cata-
lysts that would have oxygenated functional groups even after
the reduction process. These functional groups would favour
the formation of a chemical bond between GR and TiO and
2
[
14c,16b,d]
promote the dispersion of nanoparticles.
Zhang et al.
thus claimed that GR–TiO2 exhibited better photo-catalytic
degradation of methylene blue (MB) than that of CNT–TiO
2
[
14d]
[1a,10a]
with the same carbon content.
Conversely, Zhang et al.
[
17]
and Yang et al. performed studies on GR–, CNT– and C –
60
TiO products and reported that there was no advantage of
2
using GR over that of using other forms of carbon in photo-
catalysis. Evidently, the effects of different carbon materials on
TiO photo-degradation are still uncertain; thus there is a press-
2
ing need to launch a systematic study on carbon–TiO nano-
2
composites to clarify the role of carbon materials for photo-
catalysis.
Figure 1. XRD patterns of (a) the carbon precursors and (b) samples of ana-
tase with carbon materials.
Herein, we conducted a comparative study on the photo-
catalytic performance of TiO -based composites containing dif-
2
ferent types of carbon components in the degradation of rho-
damine B (RhB) dye. The AC–, GR–, CNT– and C –TiO nano-
peak located at 25.78, which corresponded to diffraction from
the (002) crystalline plane (PDF-#41-1487), whereas GO exhibit-
ed one peak located at 2q=9.428, which corresponded to the
60
2
composites were prepared with the same amount of carbon.
Our results showed that these nanocomposites showed en-
hanced photo-catalytic activity compared with that of pure
[18]
(001) plane. The oxygen functional group in GO had an en-
[18b,19]
TiO . We propose that it was the excellent adsorption proper-
larged the lattice spacing.
As illustrated by its XRD pattern
2
ties, chemical bond formation, enhanced visible-light harvest-
ing, improved dispersion and prolonged lifetime of the charge
in Figure 1a (bottom), the absence of the characteristic peak
indicated that the AC sample was amorphous. The XRD pat-
terns of the TiO and T-AC9, T-GR9, T-CNT9 and T-C 9 nano-
carriers of these carbon–TiO catalysts that promoted photo-
2
2
60
catalysis. We also found that GR, CNT and C60 could act as elec-
tron acceptors to slow down the recombination rate of elec-
tron–hole pairs, but the same was not true of AC. The three
crystalline carbon materials had essentially no differences in
their photo-catalytic performance in the degradation of RhB.
This study provides a general guideline to design carbon-
based photo-catalysts through the systematic determination of
electronic structures, optical properties and adsorption capabil-
ities.
composites are shown in Figure 1b. The diffraction patterns
could be assigned to the anatase phase (PDF-#21-1272). There
were no observable peaks from the carbon components. For
the T-GR9 sample, the (001) peak of GO might have been shift-
ed to 25.78 owing to the reduction of the oxygenated group
[14d]
during the solvothermal process,
and it overlapped with
This was a good indication
[10b,20]
the (101) peak of TiO at 25.48.
2
that GO had been reduced to GR successfully during the solvo-
thermal process. The pattern of the T-C 9 sample did not con-
60
tain any peak that could be assigned to C , because the trace
60
content of carbon might have been too small for XRD detec-
[
17]
Results and Discussion
tion, in agreement with the literature. Because the AC was
amorphous, it was reasonable to observe the sole TiO peaks
2
Phase structures and morphology
in the patterns of the T-AC9 sample. The size of the crystals in
the samples was estimated from the Scherrer formula by using
the peak located at 25.48 in the patterns of samples T, T-AC9,
Figure 1a shows the XRD patterns of the raw C , CNT, GO and
60
AC precursors before sample preparation. The sharp peaks ap-
pearing in the C60 curve (top curve) matched well with PDF-
T-GR9, T-CNT9 and T-C 9. The sizes of the corresponding crys-
60
#44-0558. The pattern of the CNT contained a typical graphite
tals in the samples were 11.8, 11.4, 11.2, 10.3 and 10.7 nm, re-
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 455

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
spectively. The above results indicated that the incorporation
of carbon materials could inhibit the aggregation of primary
particles by providing dispersive nucleation sites for TiO2,
which led to smaller particle sizes in the samples. Moreover, it
was evident that the introduction of carbon materials into TiO
2
[
13]
promoted the dispersion of the TiO nanoparticles.
2
A TEM image of a typical two-dimensional GRO thin sheet is
shown in Figure 2a. Figure 2b shows a TEM image of uniform-
ly distributed TiO particles with a size of about 12 nm; this
2
Figure 3. Raman spectra of (a) the carbon materials and (b) samples of ana-
tase with carbon materials.
Figure 2. TEM images of (a) GO, (b) T, (c) T-AC9 and (d) T-GR9. e) A SEM
image of the T-CNT9 sample. f) A high-resolution TEM (HRTEM) image of T-
corresponding two peaks of GO were located at n˜ =1356 and
C
609 sample.
ꢀ
1
1
605 cm ; these values are consistent with those reported in
[22]
the literature. The RS pattern of C featured peaks at n˜ =
60
ꢀ
1 [23]
value was consistent with calculations from the XRD results. A
1415, 1459, and 1567 cm , whereas AC was inactive in RS
(Figure 3a, bottom spectrum). For the RS patterns of the TiO2-
based composites, the results are shown in Figure 3b. Three
sharp peaks were located at n˜ =395 (B1g), 512 (A1g), and
combination of AC and TiO nanoparticles was observed, as
2
shown in Figure 2c. Owing to the crystalline structure of TiO2
and the amorphous phase of AC, the interface of the two com-
ponents could be easily distinguished, although they appeared
to have an intimate contact with each other. Figure 2d pres-
ꢀ
1
639 cm (Eg) on the left side of the RS spectra. These peaks
[16d,21,24]
were associated with the anatase crystal.
The T-GR9, T-
ents the TEM image of T-GR9, showing that the TiO nanoparti-
CNT9 and T-C 9 samples all showed two peaks located at
2
60
ꢀ1
cles were attached to a GR sheet. Figure 2e shows a SEM
about n˜ =1381 and 1595 cm , which could be assigned as the
image of TiO with CNTs. The CNTs were attached onto the sur-
graphite peak. However, compared with Raman peaks of GO,
2
face of the TiO particles. A HRTEM image of TiO –C is shown
CNT and C , remarkable shifts were observed in the composi-
2
2
60
60
[16a,25]
in Figure 2 f. The larger lattice spacing of 0.5 nm was assigned
to the (220) plane of C , whereas the smaller one of 0.352 nm
tes.
The Raman shift of three crystalline carbon materials
indicated that the synergistic effect had taken place upon the
60
was assigned to the (101) plane of crystal TiO . An intimate in-
addition of carbon materials onto TiO . It was also possible
2
2
terface was formed between C60 and TiO . These results indi-
that the TiO particles might have induced internal stresses on
2
2
[25]
cated that the incorporation of AC, GR, CNT and C60 onto TiO2
was successful and the samples obtained showed the intimate
the carbon components, which resulted in the Raman shift.
Thus, based on the RS results, we conclude that there were in-
combination of TiO and the carbon materials.
teractions between the carbon materials and the TiO phase.
2
2
The chemical states on the surface of the prepared samples
were investigated by X-ray photo-electron spectroscopy (XPS).
Figure 4a shows the C 1s XPS spectra of T, T-AC9, T-GR9, T-
CNT9 and T-C 9. The deconvolution of C 1s core-level spectra
Surface property
The surface states of the carbon materials and the fabricated
composites were elucidated by Raman spectroscopy (RS). As
shown in Figure 3a, the RS pattern of CNT contained typical D
60
demonstrated that there were four types of carbon bond in
the catalyst products. For sample T without the addition of
carbon materials, the peaks located at 284.8, 285.9, 286.7 and
ꢀ
1
ꢀ1 [20–21]
(
n˜ =1351 cm ) and G peaks ( n˜ =1582 cm ),
whereas the
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 456

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
element with high electronegativity substituted the titanium
atom, it would definitely attract the electron cloud from the ti-
tanium atom and lead to an observable shift in the XPS pat-
[5b,c]
[28]
tern.
A similar observation was reported by Chen et al.
Optical property
UV-visible spectroscopy was employed to explore the light-ab-
sorption properties of samples T, T-AC9, T-GR9, T-CNT9 and T-
C 9. The results are illustrated in Figure 5. The redshift of the
60
absorption edges in the composites was observed upon com-
parison with sample T. This phenomenon suggested that the
Figure 5. UV-visible absorption spectra of the fabricated composites.
Figure 4. XPS spectra of (a) C 1s and (b) Ti 2p of the composites.
band gap of the composites containing the carbon phase had
been narrowed. The inset in Figure 5 shows an enlargement of
the visible-light absorption spectra for the products. It was evi-
2
88.8 eV could be assigned to C=C/CꢀC, CꢀO, C=O and OꢀC= dent that there was a stronger visible-light absorption from
[
5
c
,
1
6
a
,
d
,
2
2
]
[
2
9
]
O, respectively.
The presence of carbon was probably
the samples incorporating carbon materials. Based on the
due to the adsorption of CO in air while handling the sample.
above observation, it is reasonable to claim that the chemical
interactions had been induced between the carbon materials
2
For comparison, composites with carbon components had
a new peak centred at 289.3 eV, which indicated the formation
and TiO , as already confirmed by TEM, RS and XPS results. The
2
[
5b,c,25–26]
of TiꢀOꢀC.
The intensity of the 289.3 eV peak was rela-
functional groups (CꢀO, C=O) were partly reduced during the
solvothermal process; thus leaving unpaired p electrons that
could be easily bonded to titanium. These chemical interac-
tions occurred over the surface of T-AC9, T-GR9, T-CNT9 and T-
tively weak in theT-AC9 spectrum. Even after the solvothermal
reduction, carbon–TiO composites still contained some oxy-
2
genated functional groups, which act as a bridge to connect
TiO and carbon materials. Thus, some titanium atoms might
C 9 and facilitated the formation of the TiꢀOꢀC bond and nar-
2
60
be substituted by carbon, so that the TiꢀOꢀC peak could be
rowed the band gaps of the composites, as in the case of
[
5b]
[14c,26]
observed. This could further be demonstrated by the Ti 2p
XPS spectra shown in Figure 4b. There were two main peaks
carbon doping in other studies.
This explains why the effi-
ciency of visible-light adsorption was improved in the TiO2
samples after the incorporation of AC, GR, CNT and C .
(459 and 464.7 eV) in the Ti 2p region in the spectrum of
6
0
sample T, which should be ascribed to 2p3/2 and 2p1/2 spin-or-
bital splitting photo-electrons, respectively. The splitting be-
tween the 2p3/2 and 2p1/2 core levels was about 5.7 eV; this in-
Photo-luminescence (PL) spectroscopy is known to be a pow-
erful tool to investigate the separation rate of electron–hole
pairs in semiconductors. As shown in Figure 6, the PL emission
intensities for samples T-AC9, T-GR9, T-CNT9 and T-C 9 were
4
+
[27]
dicated the predominant formation of the Ti state. The
same analysis was applied to samples T-AC9, T-GR9, T-CNT9
6
0
lower than that of pure TiO . This indicated that a slower re-
2
4
+
and T-C 9 and suggested that Ti was the overall state in the
combination rate was obtained in the samples with carbon
components. Furthermore, the emission spectra should be as-
cribed to surface defects of the composites because the wave-
length of the emission peak located at l=640–660 nm was
much longer than that of the near-band-edge emission
60
composites. A notable shift to a higher binding energy for the
Ti 2p peaks was observed upon the incorporation of carbon
materials; this indicated that the spreading of electron cloud
was reduced owing to the formation of TiꢀOꢀC. If the carbon
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 457

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
contents were prepared and examined under visible-light irra-
diation. The results showed that a large improvement in the
rate of degradation was observed upon the addition of GR.
Specifically, the removal efficiencies after irradiation for 60 min
for samples T and T-GR9 were 50 and 95%, respectively. Thus,
sample T-GR9 exhibited the optimal photo-catalytic activity.
However, the addition of excessive GR to TiO had a negative
2
effect on its catalytic performance. This could be ascribed to
the reduction of the surface of TiO exposed to light irradiatio-
2
[1a]
n.
Furthermore, optimal ratios of AC, CNT and C60 were
doped onto TiO to compare the effect of different carbon ma-
2
terials on TiO . The photo-catalytic activity followed the order
2
T-C 9>T-CNT9>T-GR9>T-AC9>T. To further evaluate the
60
Figure 6. PL spectra of the composites.
photo-catalytic performance of the composites with carbon
materials, the kinetic rate constant was determined by the
[22]
Langmuir–Hinshelwood model. For low-concentration pollu-
tants, the pseudo-first-order kinetic model can be adopted, ac-
cording to following Equations (1) and (2):
[
8f]
peak. Thus, the addition of carbon materials to TiO would
2
effectively reduce the surface defects and slow down the re-
[22,30]
combination rate of electron–hole pairs in the samples.
For sample T-AC9, the improvement in the separation rate of
electron–hole pairs might be due to the high conductivity of
AC, which favours electron transfer. Moreover, highly conduc-
tive GR, CNT and C60 could act as sinks for electrons; thus facili-
tating the separation of electrons and holes. Photo-generated
electron would be easily transported to the surface of GR, CNT
and C , and the recombination of electron–hole pairs would
dC
r ¼ ꢀ dt ¼ k
C
ð1Þ
ð2Þ
a
ꢀ
ꢁ
C0
C
ln
¼ kat
ꢀ
1
ꢀ1
60
in which r is the reaction rate (in mgL min ), C is concentra-
tion of pollutant (in mgL ), C is the adsorption equilibrium
concentration of pollutant (in mgL ), t is reaction time (in
min) and k is the apparent rate constant (in min ). Table 1
[
12,31]
ꢀ1
be hindered.
o
ꢀ
1
ꢀ
1
a
Photo-catalysis
shows the k values of catalysts T, T-AC9, T-GR9, T-CNT9 and T-
a
RhB dye is often employed in the evaluation of the perfor-
mance of catalysts for photo-catalytic degradation in water
C 9 under visible-light irradiation. It was evident that AC, GR,
60
CNT and C60 could promote TiO photo-catalytic degradation of
2
[
7]
treatment. A comparative study of dye degradation over the
fabricated composites by photo-catalysis is shown in Figure 7.
To exclude the possibility of photo-sensitisation of the dye
itself, a control experiment without catalyst was also per-
formed. The flat degradation curve (Figure 7, top curve) sug-
gested that the self-degradation of the RhB dye was negligible.
a
Table 1. Reaction rate constants (k ) for the photo-catalytic degradation
of RhB under visible-light irradiation.
[
a]
T
T-AC9
T-GR9
T-CNT9
T-C609
ꢀ1
k
R
a
2
[min
]
0.01173
0.9936
0.01897
0.9943
0.02388
0.99738
0.03558
0.9951
0.06424
0.9941
To evaluate the effect of carbon content on TiO photo-catalyt-
2
ic performance, a series of T-GR catalysts with different carbon
2
[
a] R represents the square of the correlation coefficient for kinetic linear
fitting.
RhB, whereas T-GR9, T-CNT9 and T-C 9 presented better
60
photo-catalytic performances compared with T-AC9. Mean-
while, T-C 9 provided the most rapid photo-catalytic activity,
60
followed by T-CNT9 and T-GR9 in our studies. Therefore, GR
did not seem to have an advantage in the photo-catalytic deg-
radation of RhB, compared with those of CNTs and C .
60
To explain the enhanced effects of the carbon components
on TiO photo-catalytic degradation, our interpretations were
2
based on the results of characterisation. First, the introduction
of carbon materials into TiO enhanced the adsorption ability
2
for the organic dye. As illustrated in Figure 8, the adsorption
experiment performed in the dark showed that catalysts with
carbon components had better adsorption performances than
Figure 7. Photo-catalytic degradation of RhB under visible-light irradiation
by using the fabricated composites.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 458

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
+
radical (RhBC ), the separated electron would be captured by
surface oxygen to produce reactive oxygen radicals; thereafter,
the RhB molecules would be degraded by these reactive radi-
cals. In the case of T-AC9, the incorporation of AC sped up the
charge-carrier transfer rate, which had a positive effect on
photo-catalysis. For T-GR9, T-CNT9 and T-C 9, the band gaps
60
were narrowed and could be activated under visible-light irra-
diation. In addition, GR, CNT and C60 were reported to be effec-
tive electron sinks because their corresponding work functions
[3,13,16d,22,32]
were 4.4, 4.8 and 4.3 eV,
respectively. As illustrated in
Figure 9b, a photo-excited electron both from RhB and T-GR9,
T-CNT9 and T-C 9 would be transported to the GR sheet, CNTs
60
and C , respectively, owing to their energy band structures. In
60
addition to the reactive radicals produced according to the
suggested pathway illustrated in Figure 9a, the resultant hole
Figure 8. Histogram showing the adsorption abilities of the RhB organic dye
for the samples after 15 min in the dark.
ꢀ
would also combine with surface OH or H O to form COH radi-
2
that of the same amount of pure TiO . Easy access to
2
pollutants would enhance photo-catalysis on the sur-
face of the samples. The best adsorption ability of T-
C 9 for RhB might originate from the smaller C par-
60
60
ticle size (Figure S3 in the Supporting Information)
and its good dispersion in the nanocomposite.
Second, the intimate contact between the TiO parti-
2
cles and the carbon materials would result in smaller
particle sizes and assure uniform distribution; thus
providing more active sites for photo-catalysis and
the degradation of the dye molecules. Third, chemi-
cal bond formation (TiꢀOꢀC), as verified by RS, XPS
and UV-visible spectroscopy, narrowed the band gap
of the composites, which promoted visible-light har-
vesting and generated electron–hole pairs. Fourth,
the effective separation of photo-generated elec-
tron–hole pairs was achieved upon the addition of
carbon material, as illustrated by the PL results. The
overall improved photo-catalytic performance of
carbon ꢀTiO nanocomposites compared with that of
2
bare TiO resulted from a synergistic effect of these
2
key factors.
Furthermore, photo-catalytic mechanisms were
also proposed based on the above analysis. It was re-
ported that the work function of the RhB molecule
and excited RhB were 5.45 and 3.08 eV, respective-
[
16b]
ly.
.2 eV,
The electron affinity of anatase TiO2 is
[
16d,22]
4
which is higher than that of the work
Figure 9. Proposed mechanisms for the photo-catalytic degradation of RhB over the cat-
function of excited RhB. By considering that the ab- alyst. VB=valence band.
sorption edge of samples T and T-AC9 (though its ab-
sorption edge had been redshifted) were both locat-
ed lower than l=420 nm (the cutoff filter of the irradiation
used in this study), the observed photo-catalytic activity
should be attributed to the photo-sensitisation of RhB. As
shown in Figure 9a, under visible-light irradiation, RhB could
be photo-excited to the RhB* state and the resulting electron
would be energetically favourably transferred to the conduc-
cals. Thus, it was not surprising to observe remarkably promot-
ed photo-catalytic degradation properties when using T-GR9,
T-CNT9 and T-C 9, compared with those of T and T-AC9. To
60
confirm the proposed mechanism, the COH groups were re-
corded in catalyst T-C 9 at different times. As shown in
60
Figure 10, COH was released incrementally with increasing reac-
tion time. The inset in Figure 10 exhibits the time dependence
of the PL intensity. The linear relationship indicated that the
tion band of TiO and would achieve separation of electron–
2
hole pairs. Meanwhile, if RhB* was converted into the cationic
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 459

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
five runs. Even after five runs, a negligible decrease in photo-
catalytic performance was observed; this indicated that T-C 9
6
0
was re-usable and had an excellent stability for the photo-cata-
lytic degradation of RhB.
Conclusion
AC–, GR–, CNT– and fullerene–TiO nanocomposites were suc-
2
cessfully prepared by a solvothermal method combined with
post-annealing treatment. The incorporation of carbon materi-
als onto TiO enhanced its photo-catalytic degradation of RhB
2
owing to the high adsorption ability for the dye, improved dis-
persion of active sites, enhanced visible-light harvesting with
a narrower band gap and effective separation of electron–hole
pairs. This study offers general guidelines for designing photo-
catalysts with carbon components.
Figure 10. PL spectral changes taken from basic TA solution under visible-
light irradiation over T-C609. Inset: the corresponding PL intensity measured
from the l=426 nm peak versus irradiation time.
rate of release of COH in T-C 9 was stable, which also matched Experimental Section
60
well with the kinetic analysis described in Table 1. Therefore,
the proposed mechanisms could well elucidate the photo-cata-
lytic activity of as-prepared composites. GR, CNT and C , with
Sample preparation
The raw precursors GO, AC, CNT and C60 were purchased from the
Nanjing XFNANO Materials Tech Company. GO was dispersed
evenly in anhydrous ethanol by using ultrasound to achieve a uni-
form concentration of 1 mgmL before use in sample preparation.
CNTs and C60 (as shown in Figures S2 and S3 in the Supporting In-
60
excellent adsorption abilities and conductivity, would all serve
as electron sinks in promoting TiO photo-catalytic degradation
2
ꢀ1
of RhB from the above analysis. T-AC9, without an electron
trap (AC could not serve as electron sink), showed weaker in-
formation) were both pre-treated in concentrated HNO (68%) at
3
teractions between AC and TiO , a lower rate of adsorption to
2
1308C for 4 h, whereas AC (as shown in Figure S1 in the Support-
ing Information) was treated with concentrated HNO3 (68%) at
608C for 2 h to avoid over-oxidation. Titanium(IV) n-butoxide
the dye and a larger band gap; this led to a relatively low
photo-catalytic property, compared with that of T-GR9, T-CNT9
and T-C 9. In addition, T-C 9 showed the best photo-catalytic
(
(
TBOT; 99%, Acros) was used as a precursor for titania. Acetic acid
HAc; 100%, analytic purity) served as an inhibitor to control the
60
60
performance relative to T-GR9 and T-CNT9, which could be at-
rate of hydrolysis of TBOT. Distilled water with a resistivity of
tributed to the strongest interaction of carbon and TiO , the
2
ꢀ1
1
8 MWcm was used in these experiments. Carbon–TiO compo-
2
largest adsorption capability for RhB and most effective separa-
tion of photo-induced charge carriers.
sites were synthesised by a solvothermal method followed by
a post-heating treatment. In a typical preparation of a GR–TiO2
sample, TBOT (3 mL), HAc (7 mL) and GO (3, 6, 9, or 12 mg) were
placed in anhydrous ethanol (20 mL) and mixed under magnetic
stirring for 1 h. Distilled water (5 mL) was added dropwise to the
mixture before it was transferred to a 45 mL stainless-steel auto-
clave for heating at 1808C for 10 h. The precipitate in the mixture
was collected by centrifugation (with 3000 rmp) and was washed
with anhydrous ethanol. The grey powder was dried at 808C for
In industrial applications, it is important that a catalyst is
chemically stable and can be used repeatedly. In the evaluation
of the stability of T-C 9, the sample was directly immersed
60
into pollutant solution each time without being washed after
the former run finished. Figure 11 shows the degradation effi-
ciency of T-C 9 after 60 min under visible-light irradiation for
60
6
h before calcination at 4008C for 2 h in argon. The final GR–TiO2
products, labelled as T-GR3, T-GR6, T-GR9, and T-GR12, correspond-
ed to 0.4, 0.8, 1.25 and 1.65 wt% carbon, respectively. For the
preparation of AC–, CNT– and C –TiO composites, 9 mg of AC,
60
2
CNT and C60 was used instead and samples prepared with the
same procedure as described above. The samples were referred to
as T-AC9, T-CNT9 and T-C 9, respectively. As a control, a nanocrys-
60
talline sample of TiO was also prepared and denoted by T.
2
Characterisation
XRD analysis was conducted by using a Rigaku RU300 diffractome-
ter with CuKa radiation (l=0.1540598 nm). Electron microscopy
was performed by using a Quanta F400 field-emission scanning
2
electron microscope and a Tecnai G 20S-Twin 120 kV transmission
electron microscope. The UV/Vis absorption spectra were obtained
by using a Hitachi UV/Vis spectrophotometer (U-3501). PL spectra
were recorded by using a Hitachi F-4500 fluorescence spectropho-
Figure 11. The photo-catalytic performance of catalyst T-C609 in repeated
cycles.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 460

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
tometer with an excitation wavelength of l=325 nm. XPS was per-
formed by using a PHI Model 5802 instrument with AlKa radiation
and calibrated with C 1s at 284.8 eV.
[8] a) Q. Xiang, J. Yu, B. Cheng, H. Ong, Chem. Asian J. 2010, 5, 1466–1474;
b) Y. Lu, H. Yu, S. Chen, X. Quan, H. Zhao, Environ. Sci. Technol. 2012, 46,
1
724–1730; c) Q. Zhang, D. Q. Lima, I. Lee, F. Zaera, M. Chi, Y. Yin,
Angew. Chem. 2011, 123, 7226–7230; ; Angew. Chem. Int. Ed. 2011, 50,
088–7092; d) A. Furube, L. Du, K. Hara, R. Katoh, M. Tachiya, J. Am.
7
Chem. Soc. 2007, 129, 14852–14853; e) Y. Tian, T. Tatsuma, J. Am. Chem.
Soc. 2005, 127, 7632–7637; f) Q. Deng, X. Duan, D. H. Ng, H. Tang, Y.
Yang, M. Kong, Z. Wu, W. Cai, G. Wang, ACS Appl. Mater. Interfaces 2012,
Photo-catalysis
The photo-catalytic performance of the samples was evaluated by
observing their abilities to degrade the RhB dye in aqueous solu-
tions under visible-light irradiation induced by a 300 W Xe lamp
4
, 6030–6037.
9] H. Yang, Z. Tan, Y. Liu, Z. Ma, L. Zhang, IEEE Trans. Nanotechnol. 2013,
2, 1037–1041.
[
1
(
HSX-F300, Beijing NBET Technology) with a l=420 nm cutoff
[
10] a) Y. Zhang, Z. R. Tang, X. Fu, Y. J. Xu, ACS Nano 2011, 5, 7426–7435;
b) Q. Li, H. Yang, A. Nie, X. Fan, X. Zhang, Catal. Lett. 2011, 141, 1237–
filter. The reaction chamber was placed under the light beam at
a distance of 20 cm. In a typical evaluation procedure, sample
1
9
242; c) Q. Li, H. Yang, F. Qiu, X. Zhang, J. Hazard. Mater. 2011, 192,
15–921.
(
50 mg) was placed into the RhB solution (50 mL at a concentration
ꢀ1
of 5 mgL ). The suspension was stirred for 30 min in the dark to
achieve adsorption equilibrium between RhB dye and the sample
before it was placed in the reaction chamber for visible-light expo-
sure. At a given time interval, 1.5 mL of the solution was collected
and the concentration of RhB was determined by measuring the
light absorbance of RhB at l=553 nm by means of UV/Vis spec-
troscopy. In this experiment, the RhB concentration immediately
[11] Q. Li, H. Yang, Z. Ma, X. Zhang, Catal. Commun. 2012, 17, 8–12.
[12] K. Woan, G. Pyrgiotakis, W. Sigmund, Adv. Mater. 2009, 21, 2233–2239.
13] Y.-J. Xu, Y. Zhuang, X. Fu, J. Phys. Chem. C 2010, 114, 2669–2676.
14] a) N. Yang, Y. Liu, H. Wen, Z. Tang, H. Zhao, Y. Li, D. Wang, ACS Nano
[
[
2
013, 7, 1504–1512; b) J. Zhang, Z. Zhu, Y. Tang, X. Feng, J. Mater.
Chem. A 2013, 1, 3752–3756; c) J. S. Lee, K. H. You, C. B. Park, Adv.
Mater. 2012, 24, 1133–1133; d) H. Zhang, X. Lv, Y. Li, Y. Wang, J. Li, ACS
Nano 2010, 4, 380–386; e) N. Zhang, Y. Zhang, M.-Q. Yang, Z.-R. Tang,
Y.-J. Xu, J. Catal. 2013, 299, 210–221; f) N. Zhang, M.-Q. Yang, Z.-R. Tang,
Y.-J. Xu, J. Catal. 2013, 303, 60–69; g) Y. Zhang, N. Zhang, Z.-R. Tang, Y.-
J. Xu, ACS Nano 2012, 6, 9777–9789; h) N. Zhang, Y. Zhang, Y.-J. Xu,
Nanoscale 2012, 4, 5792–5813.
after adsorption equilibrium in the dark was denoted by C . To con-
firm the presence of COH, PL spectroscopy was employed. The ap-
proach was similar to those used by Liu et al., who used tereph-
thalic acid (TA) molecules as a probe. In the analysis, (20 mg) cata-
0
[33]
ꢀ1
[
15] Q. Xiang, J. Yu, M. Jaroniec, Chem. Soc. Rev. 2012, 41, 782–796.
lyst was dispersed into a mixture (20 mL) of TA (3 mmolL ) and
ꢀ
1
[16] a) B. Cai, X. Lv, S. Gan, M. Zhou, W. Ma, T. Wu, F. Li, D. Han, L. Niu, Nano-
scale 2013, 5, 1910–1916; b) J. Zhang, Z. Xiong, X. Zhao, J. Mater.
Chem. 2011, 21, 3634–3640; c) H. Zhang, X. Fan, X. Quan, S. Chen, H.
Yu, Environ. Sci. Technol. 2011, 45, 5731–5736; d) Y. Wen, H. Ding, Y.
Shan, Nanoscale 2011, 3, 4411–4417.
NaOH (10 mmolL ) under stirring in the dark for 30 min before
visible-light irradiation. At a given time interval, 2 mL of sample
was collected to monitor the concentration of COH by measuring
the PL peak intensity of the 2-hydroxyterephthalic acid generated
at l=426 nm by using a Hitachi F-4500 fluorescence spectropho-
tometer with an excitation wavelength of l=320 nm.
[
17] M.-Q. Yang, N. Zhang, Y.-J. Xu, ACS Appl. Mater. Interfaces 2013, 5, 1156–
1164.
[
18] a) Q. Shou, J. Cheng, L. Zhang, B. J. Nelson, X. Zhang, J. Solid State
Chem. 2012, 185, 191–197; b) H. Yang, L. Zhang, X. Dong, W. Zhu, J.
Zhu, B. Nelson, X. Zhang, Nanotechnology 2012, 23, 065604–065609.
19] L. Dong, K. Shou, D. R. Frutiger, A. Subramanian, L. Zhang, B. J. Nelson,
X. Tao, X. Zhang, IEEE Trans. Nanotechnol. 2008, 7, 508–517.
20] Q. Li, X. Hou, H. Yang, Z. Ma, J. Zheng, F. Liu, X. Zhang, Z. Yuan, J. Mol.
Catal. A: Chem. 2012, 356, 121–127.
Acknowledgements
[
[
This study is supported by the Direct Grant for Research 2011/
2012 from the Faculty of Science, The Chinese University of Hong
Kong (Project code: 2060439).
[21] K. Zhou, Y. Zhu, X. Yang, X. Jiang, C. Li, New J. Chem. 2011, 35, 353–
59.
3
[
22] P. Wang, Y. Tang, Z. Dong, Z. Chen, T.-T. Lim, J. Mater. Chem. A 2013, 1,
Keywords: carbon · dyes/pigments · fullerenes · graphene ·
photochemistry · titanium
4718–4727.
[23] K. Nakamoto, M. A. McKinney, J. Chem. Educ. 2000, 77, 775–780.
24] M. Wojtoniszak, B. Zielinska, X. Chen, R. J. Kalenczuk, E. Borowiak-Palen,
J. Mater. Sci. 2012, 47, 3185–3190.
25] Y. L. Min, K. Zhang, W. Zhao, F. C. Zheng, Y. C. Chen, Y. G. Zhang, Chem.
Eng. J. 2012, 193–194, 203–210.
26] Y. Zhang, C. Pan, J. Mater. Sci. 2011, 46, 2622–2626.
27] C. Su, L. Liu, M. Zhang, Y. Zhang, C. Shao, CrystEngComm 2012, 14,
[
[
1] a) Y. Zhang, Z.-R. Tang, X. Fu, Y.-J. Xu, ACS Nano 2010, 4, 7303–7314;
b) H. B. Wu, H. H. Hng, X. W. Lou, Adv. Mater. 2012, 24, 2567–2571.
2] a) W. Liang, T. L. Church, A. T. Harris, Green Chem. 2012, 14, 968–975;
b) J. B. Joo, Q. Zhang, M. Dahl, I. Lee, J. Goebl, F. Zaera, Y. D. Yin, Energy
Environ. Sci. 2012, 5, 6321–6327; c) J. B. Joo, Q. Zhang, I. Lee, M. Dahl, F.
Zaera, Y. Yin, Adv. Funct. Mater. 2012, 22, 166–174.
[
[
[
[
3
989–3999.
28] L.-C. Chen, Y.-C. Ho, W.-S. Guo, C.-M. Huang, T.-C. Pan, Electrochim. Acta
009, 54, 3884–3891.
[
[
[
[
3] Y. Long, Y. Lu, Y. Huang, Y. Peng, Y. Lu, S.-Z. Kang, J. Mu, J. Phys. Chem. C
2
2
009, 113, 13899–13905.
4] O. Akhavan, R. Azimirad, S. Safa, M. Larijani, J. Mater. Chem. 2010, 20,
386–7392.
5] a) L. Zhang, M. S. Tse, O. K. Tan, Y. X. Wang, M. Han, J. Mater. Chem. A
013, 1, 4497–4507; b) J. Yu, G. Dai, Q. Xiang, M. Jaroniec, J. Mater.
Chem. 2011, 21, 1049–1057; c) H. Wang, Z. Wu, Y. Liu, J. Phys. Chem. C
009, 113, 13317–13324.
6] J. Zhang, C. Pan, P. Fang, J. Wei, R. Xiong, ACS Appl. Mater. Interfaces
010, 2, 1173–1176.
[
[
29] P. Gao, D. D. Sun, Chem. Asian J. 2013, 8, 2779–2786.
30] D. Wang, Z. H. Zhou, H. Yang, K. B. Shen, Y. Huang, S. Shen, J. Mater.
Chem. 2012, 22, 16306–16311.
31] R. Leary, A. Westwood, Carbon 2011, 49, 741–772.
32] A. Kongkanand, P. V. Kamat, ACS Nano 2007, 1, 13–21.
33] G. Liu, Q. Deng, H. Wang, D. H. L. Ng, M. Kong, W. Cai, G. Wang, J.
Mater. Chem. 2012, 22, 9704–9713.
7
[
[
[
2
2
[
[
2
7] W. Zhou, Z. Yin, Y. Du, X. Huang, Z. Zeng, Z. Fan, H. Liu, J. Wang, H.
Received: November 13, 2013
Zhang, Small 2013, 9, 140–147.
Published online on January 21, 2014
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 454 – 461 461