CARBON 4 9 ( 2 0 1 1 ) 5 3 1 2 –5 3 2 0

available at www.sciencedirect.com

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

Fabrication of graphene films on TiO2 nanotube arrays

for photocatalytic application

Chengbin Liu a, Yarong Teng a, Ronghua Liu a, Shenglian Luo a,*

, Yanhong Tang b,
Liuyun Chen b, Qingyun Cai a
a
State Key Laboratory of Chemo/Biosensing and Chemometrics, Hunan University, Changsha 410082, PR China
b
College of Materials Science and Engineering, Hunan University, Changsha 410082, PR China

A R T I C L E I N F O A B S T R A C T

Article history: Graphene film was formed on the surface of titanium dioxide nanotube (TiO2 NT) arrays
Received 24 March 2011 through in situ electrochemical reduction of a graphene oxide dispersion by cyclic voltam-
Accepted 29 July 2011 metry. The residual oxygen-containing groups and other structural defects such as
Available online 4 August 2011 sp3-hybridized carbons in the electrodeposited graphene were further removed by photo-
assisted reduction of the underlying TiO2 NTs, thus achieving the maximum restoration of
p-conjugation in the graphene planes. Spectroscopic, electrochemical, and photoelectro-
chemical techniques were used to characterize the graphene films, and the use of the result-
ing graphene–TiO2 NT material in photocatalysis was investigated. The results showed that
the graphene–TiO2 NT material exhibited a greatly improved photocatalytic activity com-
pared with unmodified TiO2 NTs.
 2011 Elsevier Ltd. All rights reserved.

1. Introduction a series of attributes, such as excellent electron capture and

transport properties [8,9], high transmittance [10], large spe-
Graphene, a one-atom-thick sp2-hybridized carbon sheet, has cific surface area [11], and good affinity toward organic mole-
attracted a great deal of scientific interest since its discovery cules [12,13], graphene has recently drawn considerable
by Geim and co-workers [1]. Owing to its outstanding interest in photocatalysis [14–16]. It has been demonstrated
mechanical, electrical, and thermal properties [2,3], graphene that the incorporation of graphene into TiO2 greatly facili-
holds enormous potential in a variety of applications, includ- tated effective photodegradation of organic compounds. Nev-
ing electronic devices, transparent electrodes, supercapaci- ertheless, up to now, most of these studies focus on TiO2
tors, sensors, and composites [4,5]. On the other hand, nanoparticles. Highly ordered TiO2 NT arrays fabricated by di-
semiconductor-based photocatalysts have attracted much rect anodization of titanium (Ti) offer a much larger surface
attention over the world for environmental remediation, of area and facilitate more efficient electron transport than
which TiO2 is considered as the most promising material the randomly oriented TiO2 nanoparticles [17,18], and are cur-
due to its relatively high activity, nontoxicity, chemical stabil- rently attracting a high interest in various applications
ity, and low cost, while the practical applications are still re- including photocatalysis. However, to our best knowledge,
tarded by several limitations [6,7]: (i) wide band gap (3.0 and there is no report about the graphene–TiO2 NT materials,
3.2 eV for the rutile and anatase phases, respectively), limit- which may primarily be restricted by approaches to the com-
ing the photoactivation only in the ultraviolet light region; bination of graphene with TiO2 NTs.
(ii) high carrier-recombination probability; and (iii) poor For the graphene preparation, various methods, such as
adsorbing affinity toward organic molecules. On account of mechanical exfoliation [1], epitaxial growth [19], chemical

* Corresponding author: Fax: +86 731 88823805.

E-mail address: [email protected] (S. Luo).
0008-6223/$ - see front matter  2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbon.2011.07.051
 CARBON 4 9 ( 20 1 1 ) 5 3 1 2–53 2 0 5313

vapor deposition [20], and reduction of graphene oxide (GO) ribbon was anodized at 15 V in an electrolyte containing
derived from exfoliated graphite oxide [21,22], have been 0.1 M NaF and 0.5 M NaHSO4 at room temperature for 5 h in
proposed, among which the last approach provides access a two-electrode system with a platinum cathode, resulting
to graphene-based composite materials. For example, by in the TiO2 NT arrays with a pore size of 90–100 nm, a length
chemical, photocatalytic, or hydro/solvothermal reduction of 320 nm and an efficient electrode area of 4 cm · 1 cm. Be-
of GO in the presence of TiO2 nanoparticles or Ti precursors fore the modification of graphene, the TiO2 NTs were an-
[9,14–16,23,24], powdery graphene–TiO2 composites can be nealed at 500 C for 3 h with heating and cooling rates of
easily prepared. These methods, however, seem not suitable 2 C min1 to obtain photoactive anatase phase. The anatase
for the fabrication of graphene–TiO2 NT materials. phase was not influenced by graphene electrodeposition
Recently, we have reported that GO in aqueous solutions (see Fig. S1 in Supporting Information).
can be directly electrochemically reduced using cyclic voltam-
metric method to yield graphene films on electrode surfaces 2.3. Preparation of graphene films on TiO2 NT arrays
[25,26]. The graphene electrodeposition has been proved to be
driven by the difference in solubility between the GO and the Graphite oxide was synthesized from graphite powder by the
electrochemically reduced GO in aqueous electrolytes. In the Hummers’ method [28]. The obtained graphite oxide powder
present work, by using TiO2 NTs grown on a Ti foil (TiO2 NTs/ was exfoliated in 0.1 M pH 9.18 phosphate buffer solution
Ti) as the electrode, we have prepared graphene film on TiO2 (PBS, Na2HPO4) by ultrasonication for 30 min to form a homo-
NTs to obtain novel graphene–TiO2 NT hybrid photocatalyst. geneous brown GO colloidal dispersion with a concentration
More interestingly, the residual oxygen-containing groups in of 0.5 mg mL1. The cyclic voltammetric reduction was per-
the electrodeposited graphene (EG) film [25,26] can be further formed in the GO dispersion with magnetic stirring, using a
removed by photo-assisted reduction of the underlying TiO2 three-electrode system (the annealed TiO2 NTs/Ti sheet as
NTs, achieving the maximum restoration of p-conjugation in working electrode, Pt foil as counter electrode, and saturated
the graphene plane. The electrochemical preparation and calomel electrode (SCE) as reference electrode) on a CHI 660D
photo-assisted reduction of the graphene film were illustrated electrochemical workstation (CH Instruments, Shanghai). The
in Fig. 1. The preparation, morphology and structure, light scan was from 1.5 to 1 V at a rate of 50 mV s1, and the load-
absorption, photoelectrochemical property, adsorptivity to or- ing of graphene sheets was controlled by the number of po-
ganic molecules, and photocatalytic activity of the graphene– tential cycles. After deposition, the working electrode was
TiO2 NT material were investigated in detail. washed with DI water, and then dried at room temperature.

2. Experimental 2.4. Photocurrent measurements

2.1. Chemicals and materials Photocurrent studies were performed in a photoelectrochem-

ical quartz cell containing 1.0 M KOH solution on the CHI 660D
Titanium foil (99.8%, 0.127 mm thick) was purchased from electrochemical workstation, using the three-electrode con-
Aldrich (Milwaukee, WI). Anthracene-9-carboxylic acid figuration with the TiO2 NTs/Ti or graphene modified TiO2
(9-AnCOOH) was obtained from Johnson Matthey Alfa AE- NTs/Ti (graphene–TiO2 NTs/Ti) sheet as the working electrode
SAR. Graphite powder about 50 lm was purchased from and, a Pt foil and an SCE as the counter electrode and reference
Shanghai Carbon Co., Ltd. Sodium bisulfate (NaHSO4), so- electrode, respectively. A 500 W xenon arc lamp (CHF-XQ-
dium fluoride (NaF) and all other reagents were of analytical 500W, Beijing Changtuo Co., Ltd.) served as simulated solar
grade and used as received. Deionized (DI) water was used light irradiation source with the photon flux of 100 mW cm2
for the preparation of all aqueous solutions. measured by NOVA Oriel 70260 with a thermodetector.

2.2. Preparation of TiO2 NT arrays 2.5. Adsorptivity measurements

TiO2 NT arrays were prepared according to our previous work The adsorptivity of the graphene–TiO2 NT material was inves-
[27]. Prior to anodization, Ti ribbons were in turn ultrasoni- tigated by placing graphene–TiO2 NTs/Ti electrodes in a
cally cleaned in acetone and ethanol. The cleaned titanium 1.0 mg L1 anthracene-9-carboxylic acid (9-AnCOOH) solution

Fig. 1 – Illustration for preparing the graphene–TiO2 NT hybrid material. TiO2 NT arrays were obtained by anodic oxidation of a
Ti foil (see Section 2), and then GO in a dispersion was electrochemically reduced by cyclic voltammetry to yield EG film on the
TiO2 NTs. A xenon arc lamp as simulated solar light source was applied to the EG film to obtain photoreduced EG (PEG) film.
5314 CARBON 4 9 ( 2 0 1 1 ) 5 3 1 2 –5 3 2 0

for 2 h and evaluating the remaining 9-AnCOOH concentration

by high-performance liquid chromatography (Agilent 1100
HPLC with a C18ODS column and an ultraviolet detector).

2.6. Photocatalytic degradation of 9-AnCOOH

The photocatalysis degradation experiments were carried out

by vertically placing the graphene–TiO2 NTs/Ti electrode in a
quartz beaker containing 60 mL of 10.0 mg L1 9-AnCOOH
with magnetic stirring. A 500 W xenon arc lamp served as
the light source to simulate solar light irradiation. The con-
centration of 9-AnCOOH in the solution was determined as
a function of irradiation time from the absorbance change
at a wavelength of 253 nm using a UV–vis spectrophotometer
(Cary 300, Varian, USA) whose calibration curve had been
determined in advance.

2.7. Characterization

The scanning electron microscope (SEM) images were

acquired by using a field emission SEM (FESEM, Hitachi, Model
S-4800). The fast Fourier transform (FFT) image was obtained
in a JEM-3010 high-resolution transmission electron micro-
scope. Fourier transform infrared (FTIR) reflectance spectra
were recorded on a FD-5DX FTIR spectrometer. UV–vis diffuse Fig. 2 – (a) Typical CVs of 0.5 mg mL1 GO in 0.1 M, pH 9.18,
reflectance spectra (DRS) were recorded on a Cary 300 Conc PBS on TiO2 NTs/Ti; (b) CVs of TiO2 NTs/Ti and EG-TiO2 NTs/
UV–vis spectrophotometer with an integrating sphere. Ti in PBS (0.1 M, pH 9.18) in comparison with CV of GO
electrolysis on TiO2 NTs/Ti. Scan rate: 50 mV s1.
3. Results and discussion

3.1. Electrochemical synthesis of graphene films advantage of the electrodeposition method is its easy control
over graphene loading by choosing the number of the poten-
Fig. 2a shows the cyclic voltammograms (CVs) for electro- tial cycling. Fe(CN)63/4 redox probe was selected to evaluate
chemical synthesis of graphene. Two anodic peaks (I and IV) the optimal loading amount of graphene. Fig. 3a displays the
and two cathodic peaks (II and III) were observed during the CVs of 5.0 mM K3Fe(CN)6 in a 1.0 M KCl solution on TiO2 NTs/
electrolysis of GO. Successive cyclic voltammetry scans Ti and EG-TiO2 NTs/Ti electrodes with different cycles of
showed that the current of all the peaks increased with graphene electrodeposition. At the potential range of 1.2 to
increasing the cycle number, suggesting that the GO in the dis- 1.8 V, TiO2 NTs/Ti only shows a reduction peak, while quasi-
persion was suffering from electrochemical reduction to yield reversible redox behavior of Fe(CN)63/4 is observed on the
conductive graphene that continuously deposited onto the EG-TiO2 NTs/Ti electrodes, confirming the formation of con-
TiO2 NTs/Ti electrode [25,26]. To assign the peaks, the CVs of ductive graphene by the cyclic voltammetric method, which
bare TiO2 NTs/Ti electrode and the EG modified TiO2 NTs/Ti promotes electron transfer between the Fe(CN)63/4 and the
(EG-TiO2 NTs/Ti) electrode in 0.1 M PBS (pH 9.18) were com- Ti electrode [33,34]. The oxidation peak current increases with
pared with the CV of GO electrolysis (Fig. 2b). The TiO2 NTs/ increasing cycle numbers from 0 to 26, and then drops beyond
Ti electrode itself exhibits a strong oxidation peak at a much 26 cycles (Fig. 3b). The initial increase in oxidation peak cur-
negative potential that accounts for the peak IV. The cathodic rent is attributed to the enhanced conductivity with increas-
peak III present in GO electrolysis disappeared when the EG- ing graphene loading. This is confirmed by the
TiO2 NTs/Ti electrode was swept in PBS, so the peak III should corresponding SEM images shown in Fig. 4, in which the sur-
be ascribed to the irreversible electrochemical reduction of GO face coverage of graphene clearly increases with the increased
[29–32]. Peaks I and II are presented both in the CV of GO elec- cycle numbers. At 5, 11, and 17 cycles, graphene islands are
trolysis and in the CV of EG-TiO2 NTs/Ti electrode recorded in observed to coexist with uncovered regions of TiO2 NT arrays,
PBS, and therefore they should be ascribed to the residual oxy- and at 26 cycles, the islands merge into a closed graphene film
gen-containing groups on graphene planes that are too stable layer on the TiO2 NT surface. At 32 cycles, the TiO2 NTs are
to be reduced by the cyclic voltammetric method [29,30]. hardly discernible because thicker graphene film was formed.
For the thicker films, the graphene sheets are supposed to
3.2. Optimization of graphene loading and SEM character- restack to form graphite-like structure through strong p–p
ization stacking interactions [35], thus resulting in slowish electron-
transfer rates relative to the 26-cycle EG film. The EG films
The loading amount of graphene affects the performance of are significantly smooth on the large TiO2 NT substrate (in line
the graphene–TiO2 composites [14,23,24], and a prominent with the tentative structure illustrated in Fig. 1), indicating
 CARBON 4 9 ( 20 1 1 ) 5 3 1 2–53 2 0 5315

TiO2 NTs upon each illumination, while the photocurrent of

the EG-TiO2 NT material first promptly yielded to
1.0 mA cm2, expectedly larger than 0.65 mA cm2 of the
bare TiO2 NTs, and then gradually increased with extending
irradiation time during the first light ‘‘on’’ process, indicative
of photoreaction occurring in the EG film; after about 80 s of
illumination, the photocurrent reached a steady state
(1.92 mA cm2). The significant increase of the redox peak
current of Fe(CN)63/4 and photocurrent on the EG-TiO2
NTs/Ti electrode during illumination suggests that the p-
electron conjugation within the EG sheets was further re-
stored (agreement with the good crystallinity shown by the
inset in Fig. 4f) that leaded to an improved electrical conduc-
tivity of the EG film. After the photoreduction by the first
light ‘‘on’’, the photocurrent response of the EG-TiO2 NTs/Ti
electrode kept fairly stable under subsequent repetitive light
irradiation, indicating robust adhesion between the graph-
ene film and the underlying TiO2 NTs that favors the direct
use of EG-TiO2 NTs/Ti as a whole.
FTIR reflectance spectroscopy was employed to analyze
the composition change of EG film caused by irradiation
(Fig. 5c). The band appearing at 1630 cm1 is attributed to
the skeletal C@C vibration of graphene [14,29]. The intensity
of O–H stretching band (coupling mO–H of Ti–OH or C–OH) at
3740 cm1 in EG-TiO2 NTs is stronger than that in bare TiO2
Fig. 3 – (a) CVs of TiO2 NTs/Ti and EG-TiO2 NTs/Ti with NTs, suggesting the existence of carbonaceous hydroxyl
different potential cycles (demonstrated by the digits) of groups (C–OH) in the as-prepared EG film, while after light
graphene electrodeposition in a 1.0 M KCl solution exposure for 2 min, its intensity decreases to the level of bare
containing 5.0 mM K3Fe(CN)6 at the scan rate of 50 mV s1; TiO2 NTs again, indicating that the C–OH groups were photo-
(b) magnification of the oxidation peak current marked by chemically removed in this case. Simultaneously, the inten-
the rectangle in panel (a), inset is oxidation peak current sity of C@O stretching (1732 cm1), O–H bending
with cycle numbers. (1453 cm1), and C–O stretching (1250 cm1) vibrations de-
creases a lot after 2 min of irradiation, and the CH2 stretching
vibrations at 2940 (asymmetric) and 2860 cm1 (symmetric)
ordered alignment of the EG sheets in the film [36]. The as- present in the as-prepared EG film disappear entirely. These
prepared EG sheet showed poor crystallinity (data not shown), results indicate that most of the oxygen functional groups
while after photocatalytic reduction (discussed below), the and sp3 defects in the EG film were removed by the photore-
FFT image (inset in Fig. 4f) shows well-defined sixfold symmet- duction. The removal of oxygen functional groups was fur-
ric diffraction spots, suggesting a crystallized graphene struc- ther proved by the CVs of EG-TiO2 NTs in PBS (Fig. 5d),
ture. Both the high crystallinity and the ordered alignment of where the redox pair of oxygen-containing groups (peaks I
graphene sheets in film are favorable for electron transport. and II) in the as-prepared EG film were absent after
photoreduction.
3.3. Photo-assisted reduction of EG The photoreduction mechanism of TiO2 toward GO has
been established [9,37,38]. Upon irradiation, the TiO2 yields
Since the cyclic voltammetric reduction of GO cannot thor- electron–hole pairs, and the holes are scavenged by sol-
oughly remove the oxygen-containing groups from the vents, in our case, by H2O to generate OH radicals and pro-
graphene plane, another interesting aspect is the combina- tons. The accumulated electrons within the TiO2 then
tion of photoreduction. Fig. 5a shows the photoreduction interact with the EG film and reduce its residual oxygen
process monitored with CVs of Fe(CN)63/4 on a EG-TiO2 functional groups. These procedures can be described by
NTs/Ti (26-cycle electrodeposited EG film was employed for Eqs. (1)–(3):
all the experiments in Fig. 5) electrode as a function of irra-
TiO2 þ hm ! TiO2 ðh þ eÞ ð1Þ
diation time (a 500 W xenon arc lamp at a photo flux of
TiO2 ðh þ eÞ þ H2 O ! TiO2 ðeÞ þ  OH þ Hþ ð2Þ
100 mW cm2 was used for the illumination). The redox
TiO2 e þ EG ! TiO2 þ photoreduced EG ð3Þ
peak current of Fe(CN)63/4 increased with extending irradi-
ation time, and for a certain time (within 90 s), the peak cur- As demonstrated in Fig. 5, photo-irradiation led to graph-
rent did not change anymore, suggesting the completion of ene with decreased structural defects and enhanced electron
the photochemical reduction. Coincident with the redox transportation, and therefore, in the following experiments,
peak current change, the photocurrent of EG-TiO2 NTs also photo-assisted reduction was performed to the EG films prior
increased during illumination. As shown in Fig. 5b, prompt to measurements, and the photoreduced EG films are referred
and stable photocurrent response was observed for the bare to as PEG films (demonstrated in Fig. 1).
5316 CARBON 4 9 ( 2 0 1 1 ) 5 3 1 2 –5 3 2 0

Fig. 4 – SEM images of TiO2 NT arrays (a), and TiO2 NT arrays covered by graphene electrodeposited with five cycles (b), 11
cycles (c), 17 cycles (d), 26 cycles (e) and 32 cycles (f). Partial graphene islands are marked by the dashed lines for the eye guide.
The inset is a typical FFT image from photoreduced graphene sheets.

3.4. Photocatalytic application of PEG-TiO2 NT materials the graphene introduction. Noticeably, both the light absorp-
tion and the photocurrent of PEG-TiO2 NTs reached to maxi-
As aforementioned, the practical photodegradation efficiency mal values at 26-cycle graphene electrodeposition resulting
of TiO2 is retarded by: (i) wide band gap; (ii) high carrier- from the formation of the largest interface area between
recombination probability; and (iii) poor adsorbing affinity to- TiO2 NT and graphene film (referring to the SEM images in
ward organic molecules. Next, the TiO2 NT photocatalysts Fig. 4), while further loading of graphene sheets on TiO2 NTs
modified with graphene will be evaluated in terms of light led to the decreases of the light absorptions and the photocur-
absorption, charge separation, and adsorptivity as well as rents, which can be attributed to the decline of light transmit-
photodegradation efficiency. tance through thicker graphene films to TiO2 surface [42,43].
Fig. 6a and b shows the DRS and photocurrent responses of The adsorptivities of the graphene–TiO2 NT hybrid materi-
different types of TiO2 NTs, respectively. Besides the intrinsic als were evaluated by choosing a typical derivative of persis-
band gap absorption of TiO2 at wavelengths less than 400 nm, tent organic pollutant, 9-AnCOOH. Because of the high
the bare TiO2 NTs also exhibit visible light activity (Fig. 6a), specific surface area of graphene nanosheets and their good
due to the presence of structural defects in the TiO2 NTs affinity toward organic molecules through hydrophobic and
[39,40]. In comparison with the bare TiO2 NTs, the modified specific p–p interactions [12,13], all the graphene modified
TiO2 NTs display that all the absorption edges are red-shifted TiO2 NTs indeed exhibit improved adsorptivity toward
accompanied by significantly enhanced absorption intensi- 9-AnCOOH (Fig. 6c) than the bare TiO2 NTs. Again, the maxi-
ties (Fig. 6a) and that the photocurrents remarkably increased mum adsorption (about 6 times of that on the bare TiO2
(Fig. 6b), indicating obvious band gap narrowing [14] and en- NTs) occurred on the 26–29-cycle EG modified TiO2 NTs,
hanced generation/separation efficiency of photo-induced where graphene islands were observed to merge into a closed
electrons and holes [15,37,41,42], respectively, as a result of graphene layer and resulted in a largest surface coverage for
 CARBON 4 9 ( 20 1 1 ) 5 3 1 2–53 2 0 5317

Fig. 5 – (a) CVs showing the change of EG-TiO2 NTs/Ti electrode in a 1.0 M KCl solution containing 5.0 mM K3Fe(CN)6 as a
function of illumination time; (b) photocurrent–time characteristic during light ‘‘on–off’’ cycles on TiO2 NTs/Ti and EG-TiO2
NTs/Ti electrodes; (c) FTIR reflectance spectra of TiO2 NTs and EG-TiO2 NTs with/without photoreduction treatment; (d) CVs of
EG-TiO2 NTs/Ti before and after photoreduction in 0.1 M, pH 9.18, PBS. For all the EG-TiO2 NTs/Ti electrodes, EG film was
obtained by using 26 electrodeposition cycles.

adsorption. Further loading of graphene led to decreased electrodeposition showed the highest photocatalytic rate re-
adsorptivity of 9-AnCOOH, which is presumably due to the sulting from the synergetic effects of light absorption, charge
decreased surface energy originating from the graphite-like separation, and adsorptivity. Fig. 7b shows UV–vis spectra of
structure of the multilayer graphene films. 9-AnCOOH in situ degraded on the PEG-TiO2 NTs/Ti electrode
From the investigations above, one can find that the three with 26 electrodeposition cycles of graphene, depicting that
main limitations of pure TiO2 such as wide band gap, high car- the characteristic adsorption peaks of 9-AnCOOH slightly de-
rier-recombination probability and poor adsorbing affinity creased after 20 min of adsorption, and then drastically de-
were overcome by the graphene modification. It could be en- creased and finally disappeared at about 60 min of light
visaged that the photodegradation efficiency can be enhanced irradiation. The complete mineralization of 9-AnCOOH was
in the graphene–TiO2 NT hybrid materials. Panel a of Fig. 7 confirmed by high-performance liquid chromatography (see
compares photocatalytic processes of modified and unmodi- Fig. S2 in Supporting Information).
fied TiO2 NTs toward 9-AnCOOH (10 mg L1) under simulated The photodegradation of 9-AnCOOH by the graphene–TiO2
solar light, where direct photolytic process without a photo- NT hybrid material can be explained based on the following pro-
catalyst is also present, from which a much higher removal posed mechanism. As aforementioned in Eq. (1), light irradia-
efficiency was observed to achieve through the photocatalytic tion excited the TiO2 NTs to yield electron–hole pairs. The
process than through the photolytic process. Prior to the photogenerated electrons were captured by graphene sheets
photocatalytic degradation, all the photocatalysts were incu- [8], which increases the lifetime of the photogenerated pairs.
bated in the 9-AnCOOH solutions for 20 min. Consistent with Electrons accumulated on graphene can be transferred to dis-
Fig. 6a, 9-AnCOOH adsorption occurred initially on the PEG- solved oxygen to form the superoxides, hydroxyl radicals, and
TiO2 NTs. Nevertheless, the concentration of 9-AnCOOH strik- so forth, while the photogenerated holes are apt to react with
ingly declined upon light irradiation, indicating that photoca- surface adsorbed water or hydroxyl groups to produce hydroxyl
talytic degradation played a key role in the experiments. The radicals [44]. The hydroxyl radical species is an extremely strong
PEG-TiO2 NTs were seen to exhibit greatly enhanced photode- oxidant, which together with photogenerated hole is responsi-
gradation activity relative to the bare TiO2 NTs. For example, ble for the mineralization of organic molecules [15,45,46].
for 60 min of illumination, 9-AnCOOH was degraded with To be noted that graphene has been shown to be oxida-
the removal ratio of 46% on the bare TiO2 NTs, much lower tively degraded by the photogenerated holes of TiO2 during
than 58%, 98%, and 82% on the PEG-TiO2 NT materials with a long-time of irradiation [47], and herein, the stability of
11-cycle, 26-cycle, and 32-cycle graphene electrodeposition, the graphene–TiO2 NT photocatalyst was evaluated by con-
respectively. The PEG-TiO2 NTs with 26 cycles of graphene secutive 10 runs, where the photocatalyst sample and the
5318 CARBON 4 9 ( 2 0 1 1 ) 5 3 1 2 –5 3 2 0

Fig. 7 – (a) Irradiation time-degradation rate of 9-AnCOOH

during direct photolytic (DP) process without a photocatalyst
and photocatalytic (PC) processes on TiO2 NTs/Ti and
PEG-TiO2 NTs/Ti electrodes with different electrodeposition
cycles of graphene. The minus sign () in the time axis
denotes the duration of dark adsorption; (b) in situ UV–vis
determination of 9-AnCOOH adsorption and photodegrada
tion on the PEG-TiO2 NTs/Ti electrode with 26-cycle
Fig. 6 – (a) DRS, (b) photocurrent responses without bias electrodeposition of graphene.
potential, (c) adsorptivities toward 9-AnCOOH with an initial
concentration of 1.0 mg L1 for TiO2 NTs and PEG-TiO2 NTs.
The digits indicate the cycle numbers for graphene
electrodeposition on the TiO2 NTs.

reaction conditions were the same as those of Fig. 7b. The re-
sults in Fig. 8 indicate that the graphene–TiO2 NT catalyst was
photocatalytic stable for the degradation. After 60 min of irra-
diation in the first run, 98% of the 9-AnCOOH was removed,
and the removal ratio was still as high as 92% in the tenth
run. The slight decrease of the removal efficiency in the con-
secutive runs was probably caused by the partial decomposi-
tion of graphene. Fig. 8 – Stability of 26-cycle electrodeposited graphene
modified TiO2 NTs. For each run, 60 min of irradiation was
4. Conclusions applied to a 60 mL of 9-AnCOOH solution with the initial
concentration of 10.0 mg L1.
We have demonstrated, for the first time, the fabrication of
graphene–TiO2 NT hybrid photocatalysts by simply electrode-
positing graphene from a GO dispersion onto TiO2 NT arrays. graphene–TiO2 NT hybrid materials are dependent on the
Photo-assisted reduction by the underlying TiO2 NTs was graphene loading, the incorporation of graphene into TiO2
found to further improve the crystallinity of the EG film, NTs improved the adsorptivity toward pollutants, enhanced
and thus the charge transport of the graphene. The graph- the light absorption intensity and light absorption range,
ene–TiO2 NT materials with different graphene loading were and boosted the transport and separation of photogenerated
prepared and compared. Although the performances of the charge carriers of TiO2 NTs. Therefore, all the resulting
 CARBON 4 9 ( 20 1 1 ) 5 3 1 2–53 2 0 5319

graphene–TiO2 NT materials exhibited significantly enhanced [14] Zhang H, Lv XJ, Li YM, Wang Y, Li JH. P25–graphene composite as
photocatalytic activities compared with the bare TiO2 NTs, a high performance photocatalyst. ACS Nano 2010;4(1):380–6.
revealing a promising application of graphene in environ- [15] Ng YH, Lightcap IV, Govdwin K, Matsumura M, Kamat PV. To
what extent do graphene scaffolds improve the photovoltaic
mental systems. This study provides a simple route toward
and photocatalytic response of TiO2 nanostructured films. J
the efficient combination of graphene, or further graphene- Phys Chem Lett 2010;1(15):2222–7.
based composites, with TiO2 NTs to construct advanced hy- [16] Akhavan O, Ghaderi E. Photocatalytic reduction of graphene
brid materials for photocatalysis as well as for other various oxide nanosheets on TiO2 thin film for photoinactivation of
applications. bacteria in solar light irradiation. J Phys Chem C
2009;113(47):20214–20.
[17] Macak JM, Tsuchiya H, Schmuki P. High-aspect-ratio TiO2
Acknowledgments nanotubes by anodization of titanium. Angew Chem Int Ed
2005;44(14):2100–2.
The research was financially supported by the National Natu- [18] Roy P, Berger S, Schmuki P. TiO2 nanotubes: synthesis and
ral Science Foundation of China (No. 51078129), the National applications. Angew Chem Int Ed 2011;50(13):2904–39.
Science Foundation for Distinguished Young Scholars (No. [19] Emtsev KV, Bostwick A, Horn K, Jobst J, Kellogg GL, Ley L, et al.
Towards wafer-size graphene layers by atmospheric pressure
50725825), and the National Basic Research Program of China
graphitization of silicon carbide. Nat Mater 2009;8(3):203–7.
(No. 2009CB421601).
[20] Reina A, Jia X, Ho J, Nezich D, Son H, Bulovic V, et al. Large
area, few-layer graphene films on arbitrary substrates by
Appendix A. Supplementary data chemical vapor deposition. Nano Lett 2009;9(1):30–5.
[21] Stankovich S, Dikin DA, Piner RD, Kohlhaas KA,
Supplementary data associated with this article can be found, Kleinhammes A, Jia Y, et al. Synthesis of graphene-based
in the online version, at doi:10.1016/j.carbon.2011.07.051. nanosheets via chemical reduction of exfoliated graphite
oxide. Carbon 2007;45(7):1558–65.
[22] McAllister MJ, Li JL, Adamson DH, Schniepp HC, Abdala AA,
Liu J, et al. Single sheet functionalized graphene by oxidation
R E F E R E N C E S
and thermal expansion of graphite. Chem Mater
2007;19(18):4396–404.
[23] Shen JF, Yan B, Shi M, Ma HW, Li N, Ye MX. One step
[1] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, hydrothermal synthesis of TiO2-reduced graphene oxide
Dubonos SV, et al. Electric field effect in atomically thin sheets. J Mater Chem 2011;21(10):3415–21.
carbon films. Science 2004;306(5696):666–9. [24] Liang YY, Wang HL, Casalongue HS, Chen Z, Dai HJ. TiO2
[2] Katsnelson MI. Graphene: carbon in two dimensions. Mater nanocrystals grown on graphene as advanced photocatalytic
Today 2007;10(1–2):20–7. hybrid materials. Nano Res 2010;3(10):701–5.
[3] Geim AK. Graphene: status and prospects. Science [25] Chen LY, Tang YH, Wang K, Liu CB, Luo SL. Direct
2009;324(5934):1530–4. electrodeposition of reduced graphene oxide on glassy
[4] Schedin F, Geim AK, Morozov SV, Hill EM, Blake P, Katsnelson carbon electrode and its electrochemical application.
MI, et al. Detection of individual gas molecules adsorbed on Electrochem Commun 2011;13(2):133–7.
graphene. Nat Mater 2007;6(9):652–5. [26] Liu CB, Wang K, Luo SL, Tang YH, Chen LY. Direct
[5] Kim KS, Zhao Y, Jang H, Lee SY, Kim JM, Kim KS, et al. Large- electrodeposition of graphene enabling the one-step
scale pattern growth of graphene films for stretchable synthesis of graphene–metal nanocomposite films. Small
transparent electrodes. Nature 2009;457(7230):706–10. 2011;7(9):1203–6.
[6] Asaki R, Morikawa T, Ohwaki T, Aoki K, Taga Y. Visible-light [27] Yang LX, Cai QY. Size-controllable fabrication of noble metal
photocatalysis in nitrogen-doped titanium oxides. Science nanonets using a TiO2 template. Inorg Chem
2001;293(5528):269–71. 2006;45(24):9616–8.
[7] Chen XB, Mao SS. Titanium dioxide nanomaterials: synthesis, [28] William S, Hummers JR, Richard EO. Preparation of graphitic
properties, modifications, and applications. Chem Rev oxide. J Am Chem Soc 1958;80(6):1339.
2007;107(7):2891–959. [29] Guo HL, Wang XF, Qian QY, Wang FB, Xia XH. A green
[8] Lightcap IV, Kosel TH, Kamat PV. Anchoring semiconductor approach to the synthesis of graphene nanosheets. ACS
and metal nanoparticles on a two-dimensional catalyst mat. Nano 2009;3(9):2653–9.
Storing and shuttling electrons with reduced graphene oxide. [30] Shao YY, Wang J, Engelhard M, Wang CM, Lin YM. Facile and
Nano Lett 2010;10(2):577–83. controllable electrochemical reduction of graphene oxide
[9] Williams G, Seger B, Kamat PV. TiO2–graphene and its applications. J Mater Chem 2010;20(4):743–8.
nanocomposites. UV-assisted photocatalytic reduction of [31] Ramesha GK, Sampath S. Electrochemical reduction of
graphene oxide. ACS Nano 2008;2(7):1487–91. oriented graphene oxide films: an in situ Raman
[10] Nair RR, Blake P, Grigorenko AN, Novoselov KS, Booth TJ, spectroelectrochemical study. J Phys Chem C
Stauber T, et al. Fine structure constant defines visual 2009;113(19):7985–9.
transparency of graphene. Science 2008;320(5881):1308. [32] Zhou M, Wang YL, Zhai YM, Zhai JF, Ren W, Wang F, et al.
[11] Stoller MD, Park S, Zhu Y, An J, Ruoff RS. Graphene-based Controlled synthesis of large-area and patterned
ultracapacitors. Nano Lett 2008;8(10):3498–502. electrochemically reduced graphene oxide films. Chem Eur J
[12] Englert JM, RÖhrl J, Schmidt CD, Graupner R, Hundhausen M, 2009;15(25):6116–20.
Hauke F, et al. Soluble graphene: generation of aqueous [33] Tang LH, Wang Y, Li YM, Feng HD, Lu J, Li JH. Preparation,
graphene solutions aided by a perylenebisimide-based structure, and electrochemical properties of reduced
bolaamphiphile. Adv Mater 2009;21(42):4265–9. graphene sheet films. Adv Funct Mater 2009;19(17):2782–9.
[13] Xu YX, Bai H, Lu GW, Li C, Shi GQ. Flexible graphene films via [34] Zhou M, Zhou YM, Dong SJ. Electrochemical sensing and
the filtration of water-soluble noncovalent functionalized biosensing platform based on chemically reduced graphene
graphene sheets. J Am Chem Soc 2008;130(18):5856–7. oxide. Anal Chem 2009;81(14):5603–13.
5320 CARBON 4 9 ( 2 0 1 1 ) 5 3 1 2 –5 3 2 0

[35] Eda G, Fanchini G, Chhowalla M. Large-area ultrathin films of [41] Yang NL, Zhai J, Wang D, Chen YS, Jiang L. Two-dimensional
reduced graphene oxide as a transparent and flexible graphene bridges enhanced photoinduced charge transport
electronic material. Nat Nanotechnol 2008;3(5):270–4. in dye-sensitized solar cells. ACS Nano 2010;4(2):887–94.
[36] Lee V, Whittaker L, Jaye C, Baroudi KM, Fischer DA, Banerjee [42] Tang YB, Lee CS, Xu J, Liu ZT, Chen ZH, He ZB, et al.
S. Large-area chemically modified graphene films: Incorporation of graphenes in nanostructured TiO2 films via
electrophoretic deposition and characterization by soft molecular grafting for dye-sensitized solar cell application.
X-ray absorption spectroscopy. Chem Mater 2009;21(16): ACS Nano 2010;4(6):3482–8.
3905–16. [43] Li XL, Zhang GY, Bai XD, Sun XM, Wang XR, Wang EG, et al.
[37] Manga KK, Zhou Y, Yan YL, Loh KP. Multilayer hybrid films Highly conducting graphene sheets and Langmuir–Blodgett
consisting of alternating graphene and titania nanosheets films. Nat Nanotechnol 2008;3(9):538–42.
with ultrafast electron transfer and photoconversion [44] Linsebigler AL, Lu GQ, Yates Jr JT. Photocatalysis on TiOn
properties. Adv Funct Mater 2009;19(22):3638–43. surfaces: principles, mechanisms, and selected results.
[38] Li B, Zhang XT, Li XH, Wang L, Han RY, Liu BB, et al. Photo- Chem Rev 1995;95(3):735–58.
assisted preparation and patterning of large-area reduced [45] Liu YT, Liu RH, Liu CB, Luo SL, Yang LX, Sui F, et al. Enhanced
graphene oxide–TiO2 conductive thin film. Chem Commun photocatalysis on TiO2 nanotube arrays modified with
2010;46(20):3499–501. molecularly imprinted TiO2 thin film. J Hazard Mater
[39] Yang LX, Luo SL, Liu RH, Cai QY, Xiao Y, et al. Fabrication of 2010;182(1–3):912–8.
CdSe nanoparticles sensitized long TiO2 nanotube arrays for [46] Lu G, Niu P, Sun CH, Smith SC, Chen ZG, Lu GQ, et al. Unique
photocatalytic degradation of anthracene-9-carboxylic acid electronic structure induced high photoreactivity of sulfur-
under green monochromatic light. J Phys Chem C doped graphitic C3N4. J Am Chem Soc 2010;132(33):11642–8.
2010;114(11):4783–9. [47] Akhavan O, Abdolahad M, Esfandiar A, Mohatashamifar M.
[40] Zhuang HF, Lin CJ, Lai YK, Sun L, Li J. Some critical structure Photodegradation of graphene oxide sheets by TiO2
factors of titanium oxide nanotube array in its photocatalytic nanoparticles after a photocatalytic reduction. J Phys Chem C
activity. Environ Sci Technol 2007;41(13):4735–40. 2010;114(30):12955–9.