Research Article

www.acsami.org

Photocatalytic Synthesis of TiO2 and Reduced Graphene Oxide

Nanocomposite for Lithium Ion Battery
Jingxia Qiu, Peng Zhang, Min Ling, Sheng Li, Porun Liu, Huijun Zhao, and Shanqing Zhang*
Centre for Clean Environment and Energy, Environmental Futures Centre, and Griffith School of Environment, Gold Coast Campus,
Griffith University QLD 4222, Australia

ABSTRACT: In this work, we synthesized graphene oxide (GO) using the improved
Hummers’ oxidation method. TiO2 nanoparticles can be anchored on the GO sheets via
the abundant oxygen-containing functional groups such as epoxy, hydroxyl, carbonyl, and
carboxyl groups on the GO sheets. Using the TiO2 photocatalyst, the GO was
photocatalytically reduced under UV illumination, leading to the production of TiO2-
reduced graphene oxide (TiO2-RGO) nanocomposite. The as-prepared TiO2, TiO2-GO,
and TiO2-RGO nanocomposite were used to fabricate lithium ion batteries (LIBs) as the
active anode materials and their corresponding lithium ion insertion/extraction
performance was evaluated. The resultant LIBs of the TiO2-RGO nanocomposite
possesses more stable cyclic performance, larger reversible capacity, and better rate
capability, compared with that of the pure TiO2 and TiO2-GO samples. The
electrochemical and materials characterization suggest that the graphene network provides
efficient pathways for electron transfer, and the TiO2 nanoparticles prevent the restacking
of the graphene nanosheets, resulting in the improvement in both electric conductivity and specific capacity, respectively. This
work suggests that the TiO2 based photocatalytic method could be a simple, low-cost, and efficient approach for large-scale
production of anode materials for lithium ion batteries.
KEYWORDS: lithium ion batteries, UV photocatalysis, TiO2, reduced graphene oxide

1. INTRODUCTION such as TiO2,15,18 SnO2,19,20 Co3O4,21 and Fe3O422 anchored

Lithium ion batteries (LIBs) have been the dominant power on the graphene sheets could potentially resolve the problem of
source for a myriad of types of portable electronic devices1 and graphene stacking.
potential power source for electric vehicles.2 To enhance LIB The preparation of graphene using chemical method
performance, a variety of nanomaterials, including TiO2,3,4 commonly involves the oxidation of graphite to produce
graphene,5−7 and its composite,8,9 has been researched as graphene oxide (GO) and reduction of GO. Chemical
alternative electrode materials. oxidation method (e.g, the traditional Hummers’ method) is
TiO2 has been particularly attractive for large-scale energy probably the most economical way to prepare GO in large
storage, because TiO2 is an abundant, inexpensive, and quantity from natural graphite.23 Recently, an improved
environmentally benign material with various types of Hummers’ method was proposed by Tour’s group24 for the
nanostructures available (e.g., nanotubes,10,11 nanowires,12,13 large-scale production of GO. The proposed method shows
and nanosheets14). Also, TiO2 is structurally stable with a small significant advantages over the Hummers’ method: it has much
volume change (<4%) during Li ion insertion/extraction higher production yield, does not involve a large exothermal
processes and intrinsically safe by avoiding Li electrochemical reaction or the production of toxic gas, produces a higher
deposition.15 However, low electronic conductivity within the fraction of well-oxidized hydrophilic carbon material, and
TiO2 network and relatively low theoretic capacity are still the possesses a more regular structure.
main obstacles for their practical applications. In this regard, During the chemical oxidation of graphite, oxygen-containing
nanocomposite of TiO2 and nanostructured carbon materials functional groups, such as epoxy, hydroxyl, carbonyl, and
may be an efficient way to tackle the problems.16,17 carboxyl, are introduced onto the graphene sheets, which
Among various nanostructured carbon materials, graphene is interrupts graphitic lattice and reduces the conductivity.25 In
expected to be an excellent supporting material of the order to restore the conductivity of graphene, these functional
nanocomposite as anodes, because of its superior electronic groups are removed by chemical reduction methods, such as
conductivity, high surface-to-volume ratio, ultrathin thickness, hydrazine,26 UV-induced reduction25 or thermal reduction,26
structural flexibility, and chemical stability.2 The graphene sonolysis.27
network provides efficient pathways for electron transfer.
However, the layered graphene sheets can naturally stack into Received: April 23, 2012
multilayers and thus lose the advantages of high surface area Accepted: June 27, 2012
and intrinsic chemical and physical properties. Nanoparticles Published: June 27, 2012
Published 2012 by the American Chemical
Society 3636 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642
ACS Applied Materials & Interfaces Research Article

Kamat and co-workers reported that GO can be photo-

catalytically reduced by the TiO2 nanoparticles under the UV
illumination.25 Since the functional groups on the graphene
sheet cannot be removed completely, the reduction product of
GO is commonly referred as reduced graphene oxide (RGO).
This method has the following advantages: (1) the extend of
the photocatalytic reduction can be tuned on demand by
controlling reaction time; (2) no toxic chemicals are involved in
the reduction reaction; and, most importantly, (3) TiO2
particles can be readily dispersed and strongly bound onto
the newly RGO sheets, forming TiO2-RGO nanocomposite due
to the remaining functional groups. Rose and co-workers28
demonstrated that the photocurrent generated by the TiO2-
RGO nanocomposite electrode was enhanced remarkably (10-
fold), compared to the pure TiO2 electrode, because of the
enhanced conductivity.
In this work, inspired by the above findings, in order to
improve both electric conductivity and specific capacity, we
attempt to incorporate the TiO2 nanoparticles with graphene
nanosheets without restacking of the graphene nanosheets. We
synthesize GO using the improved Hummers’ oxidation
method and prepare the TiO2-RGO nanocomposite via the Figure 1. (a) Schematic diagram of the preparation of the TiO2-RGO
nanocomposite; (b) photographs of the products at different synthesis
UV-induced photocatalytic reduction route to reduce GO and
stages.
anchor the TiO2 nanoparticles onto the RGO sheets
simultaneously. The as-prepared nanocomposites were used
2.2. Materials Characterization. The microstructure and
for fabrication of electrodes for LIBs. It is the first time, to the morphology of TiO2-RGO were examined using transmission electron
best of our knowledge, UV-induced reduction method is microscopy (TEM) (FEI Model Tecnai 20) with an acceleration
applied to prepare the TiO2-RGO nanocomposite as electrode voltage of 200 kV. XRD patterns were obtained on X-ray diffraction
material for LIBs. The performance of the resultant LIBs was (XRD) (Model LabX-6000, Shimadzu, Japan) using CuKα radiation at
evaluated systematically. 40 kV and 40 mA over the 2θ range of 5−70 °C. The Raman spectra
of TiO2-RGO and GO were obtained using Raman spectroscopy
(Jobin Yvon Model HR800).
2. EXPERIMENTAL SECTION 2.3. Electrode Preparation. Active materials (i.e., the pure
2.1. Synthesis Procedures. a. Preparation of GO. GO was anatase TiO2, the TiO2-GO mixture, and TiO2-RGO nanocomposite)
synthesized via the oxidation of graphite using the improved were mixed with 10 wt % carbon black and 10 wt % poly(vinylidene
Hummers’ method, the oxidation method proposed by Tour’s fluoride) (PVDF, MW ≈ 534 000, Aldrich) in N-methyl-2-pyrrolidone
group.24 Briefly, 2 g graphite and 12 g KMnO4 were placed into a (NMP, 99.5%, Aldrich) solvent to form a homogeneous slurry,
mixture of H2SO4/H3PO4 (240:26.7 mL) under vigorous stirring in respectively. The resultant slurries were uniformly coated onto Cu foils
sequence. The resultant solution was continually stirred at 50 °C for with an area of ca. 1 cm2. The weight of the active material is ca. 1−2
12 h. After the solution was cooled to room temperature, it was poured mg. All the pasted Cu foils, i.e., the anodes, were dried in a vacuum
onto ice (ca. 280 mL) and mixed with 3 mL of 30% H2O2. The oven at 120 °C for 12 h and pressed using a double-roll compressor.
solution was centrifuged and washed with 30% HCl, deionized water Using the anodes, CR2032 coin-type cells were assembled in an
and ethanol in order, and filtered with a 0.45-μm PTFE membrane. argon-filled M-Braun glovebox. A porous polypropylene film was used
The residue on the filter (i.e., GO) was dried overnight at room as the separator, a lithium sheet as the counter electrode, and 1 M
temperature. LiPF6 in a 1:1 (w/w) mixture of ethylene carbonate (EC) and
b. Preparation of Anatase TiO2. Anatase TiO2 nanoparticles were dimethyl carbonate (DMC) as the electrolyte. To measure the
prepared according to our previous work.29 Briefly, a mixture of 25 mL electrochemical capacity and cycle life of the working electrodes, the
of titanium butoxide and 8 mL of propan-2-ol was added dropwisely to cells were galvanostatically charged and discharged using LAND-
300 mL of a 0.1 M nitric acid solution under vigorous stirring, CT2001A battery tester (Wuhan, PRC) in a voltage range from 0.01
resulting in a slurry. The slurry was heated to 80 °C and stirred to 3.0 V vs Li/Li+ at the current densities of 100, 200, 400, 800, and
vigorously for 10 h to achieve peptization. The resulting colloid was 1600 mA/g, respectively.
then hydrothermally treated in an autoclave at 200 °C for 12 h. The Cyclic voltammograms (CVs) and electrochemical impedance
colloidal suspension was then introduced into a rotary evaporator and spectroscopy (EIS) were performed using a CHI 660D electro-
evaporated to a final solid concentration of ca. 2% w/v with particle chemical workstation (CH Instrument, Shanghai, PRC). CVs were
sizes ranging from 8 nm to 10 nm. recorded between 3.0 V and 0.0 V at a scan rate of 0.1 mV/s, using the
c. Preparation of the TiO2-RGO Nanocomposite. As shown in composite as the working electrode and a lithium sheet as both
Figure 1, a homogeneous TiO2-GO suspension was first obtained by counter electrode and reference electrode. AC impendence spectra
mixing a 0.04 g of GO in ethanol and 25 mL of anatase TiO2 solution were carried out with amplitude of 10 mV over the frequency range
(ca. 2% w/v) in an ultrasonic bath. The sample was then placed into a from 100 kHz to 0.01 Hz.
photoreaction station under the UV irradiation of a 1000-W xenon
lamp and continuous stirring for 2−5 h. The infrared light was blocked 3. RESULTS AND DISCUSSION
by a UV-band-pass filter. The resultant solution were filtered and
rinsed with deionized water three times. Finally, the residue solids on 3.1. The Synthesis of TiO2-RGO. The GO used to prepare
the filter membrane, i.e., TiO2-RGO nanocomposite, were dried in a the TiO2-RGO composite in this study was synthesized using
vacuum oven at 60 °C for 12 h. Part of the TiO2-GO suspension was the improved GO synthesis method.24 In contrast to the
dried and assembled in coin cells for comparison purposes, using the traditional Hummers’ method, this method can generate a
same procedures as the TiO2-RGO. larger amount of hydrophilic oxidized graphene material with
3637 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642
ACS Applied Materials & Interfaces Research Article

high production yield. We obtained 1.8 g of GO product from

1.0 g of graphite, which is close to the reported production
yield.24 Moreover, the reaction can be easily controlled and
does not generate toxic gas.
The TiO2 colloid was freshly prepared by the hydrolysis of
titanium butoxide.29 It has a milky white color. The process of
the UV induced reduction reaction is shown in Figure 1. When
the TiO2 solution is ultrasonicated with GO, a stable and
homogeneous TiO2-GO mixture is obtained via the afore-
mentioned interactions. The mixture has a light milky brown
color (see Figure 1b). When the TiO2-GO colloid is exposed to
the UV light, the charge separation takes place and produces
electron−hole pairs in TiO2 nanoparticles. The photoholes
could easily capture the electrons from organic compounds (i.e,
ethanol in this case). Under UV illumination, the electrons are
continuously promoted to the conduction band at the surface
of TiO2 surface.30 The electrons are able to reduce the
functional groups, including epoxy, hydroxyl, carbonyl, and
carboxylic functional groups on the surface of GO sheets.25,31
As a result, GO is converted into RGO, which is evidenced by
the fact that the solution color changed from light brown (the
color of TiO2-GO) into black (the color of TiO2-RGO) as
shown in Figure 1b.25,28 This anchoring of TiO2 nanoparticles
onto RGO sheets could prevent the TiO2 nanoparticles from
agglomeration and could minimize the restacking of RGO
sheets, which will be beneficial to maintaining high surface areas
and the stability of the nanocomposite during lithium
insertion/extraction.
3.2. Materials Characterization. The structural composi- Figure 2. (a) XRD patterns of the graphite, GO, RGO, TiO2, and
tion of the graphite, as-prepared GO, TiO2 nanoparticles, and TiO2-RGO composites. (b) Raman spectra of the as-prepared GO and
TiO2-RGO nanocomposite were examined using XRD (Figure TiO2-RGO.
2a). Before the oxidation reaction, the graphite showed a sharp
and strong diffraction peak at 2θ = 26.4° (see Figure 2a). After motion of symmetric sp2 C−C bond.34 Compared with the
the oxidation reaction, the resultant GO shows a distinctive intensity ratio of ID/IG of GO (1.06), the ratio of the
(002) peak at 2θ = 10.9°; this illustrates that most of the nanocomposite increased to 1.13, suggesting the decrease of
natural graphite was oxidized to GO and the d-spacing was the oxygen-functional groups on GO.35 For TiO2, due to Eg
expanded from 3.4 Å to 9.1 Å.24,32 The TiO2-RGO composites vibration modes of anatase, a sharp Raman scattering peak for
exhibits similar diffraction peaks corresponding to (101), (004), free-standing anatase nanoparticles is commonly observed at
(200), (105), and (211) of pure anatase TiO2 particles. The 143 cm−1. In contrast, the peak was observed at 153 cm−1. The
diffraction peak for RGO was reported to appear at a 2θ angle blue shift from 143 cm−1 to 153 cm−1 may be attributed to the
of ∼24.57° (d-spacing ≈ 3.62 Å) and ∼23° (d-spacing ≈ 3.86 interaction of RGO and TiO2.36 The peaks at ∼195, 400, 516,
Å).33 Interestingly, the XRD peaks corresponding to GO or and 639 cm−1 correspond to the different vibration modes of
RGO in the composite cannot be observed in Figure 2a. This anatase.37 Overall, the Raman spectra suggest that the
phenomenon was attributed to the insufficient amount of nanocomposite contains RGO and TiO2, which is consistent
carbon and relatively low diffraction intensity of the RGO in with the XRD results.
the literature.18 In this work, this is more likely because the To investigate the morphology and structure of the products,
anchored TiO2 nanoparticles prevent the graphene sheets from TEM and SEM images of the TiO2-RGO composite were
restacking, substantially expand the d-spacing of the graphene obtained as in Figure 3. The high-resolution TEM (HRTEM)
layers, disrupt the layered and ordered structure, and image (see Figure 3b) indicates that the crystal lattice fringes
consequently lead to the disappearance of the reflection with d-spacing of 0.352 nm corresponding to the (101) plane of
peaks. This result also suggests the RGO sheets are effectively the anatase TiO2, which is consistent with the XRD results. The
separated and highly dispersed into the TiO2 matrix. crystal lattice fringes of the RGO could not be observed by
In order to confirm the chemical composition of the HRTEM, possibly because of their ultrathin structure.
composite, the Raman spectroscopy was used to characterize Figure 3a shows that the TiO2 nanoparticles are well-
the RGO/TiO2 nanocomposite as well as the as-prepared GO dispersed and uniformly anchored on the surface of the RGO
(Figure 2b). According to the results examined by the sheet. The SEM image (Figure 3c) of the TiO2-RGO sample
thermogravimetric analysis (TGA), the as-prepared composite demonstrates the homogeneity of the composite in a large
contains 5.6% (w/w) of RGO. In Raman spectra, both samples scale. In particular, only a small portion of RGO sheets and
exhibit two strong peaks, denoted as the disorder peak (D, trace amount of freestanding TiO2 can be observed, while most
centered at 1350 cm−1) and the graphitic peak (G, at ∼1580 of the RGO sheets are covered by the TiO2 nanoparticles (i.e.,
cm−1). The D peak suggests the disruption of the symmetrical RGO-bonded TiO2). Also, majority of the TiO2 particles are
hexagonal graphitic lattice, which associates with internal anchored on the RGO sheets (with very small amount of free-
structure; and the G peak indicates the in-plane stretching standing TiO2 nanoparticles), which is very important, because
3638 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642
ACS Applied Materials & Interfaces Research Article

plane, while our TiO2 nanoparticles are scattered evenly on the

RGO plane (see Figures 3a and 3c).
3.3. Evaluation of Battery Performance. It is well-
established that the lithium ion insertion/extraction processes
at anatase TiO2 nanoparticles proceed according to a reversible
reaction:
TiO2 + x Li+ + x e− ↔ LixTiO2 (1)
The maximum value of x for a reversible reaction at room
temperature is 0.5,39 which corresponds to a capacity of 168
mAh/g. Usually, the voltage window for TiO2-based anode
materials is 1−3 V, because RGO contributes charge and
discharge capacity under 1 V; we set the voltage window
between 0.01 V to 3 V to obtain the full capacity of the
composite. Figure 4 presents the charge/discharge profiles of

Figure 3. (a) TEM, (b) HRTEM, and (c) SEM images of the TiO2-
RGO nanocomposite.

this ensures the efficient electron collection via the RGO sheets
during the insertion/extraction processes.
The formation mechanism of the TiO2-RGO composite is
proposed as follows. GO possesses oxygen-containing func- Figure 4. Galvanostatic charge−discharge curves of the TiO2, TiO2-
GO, and TiO2-RGO cycled at 1st, 2nd, 10th, and 30th between 0.1 V
tional groups including epoxy, hydroxyl, carbonyl, and carboxyl
and 3.0 V (vs Li/Li+) at a current density of 100 mA/g.
functional groups. Epoxy and hydroxyl functional groups are
the major part, existing on the basal plane of GO sheets, while
carboxyl and carbonyl are the minor part, occurring on the edge the TiO2, TiO2-GO, and TiO2-RGO composite electrode of the
of GO sheets.33,38 Because TiO2 nanoparticles and its surface 1st, 2nd, 10th, and 30th cycles at a current density of 100 mA/
hydroxyl groups can readily interact with these functional g. The discharging plateaus at ∼1.75 V (insertion process)
groups, TiO2 nanoparticles can distribute uniformly on the indicates the Li-intercalation into anatase TiO2 lattice and the
surface of GO, which also facilitates the photocatalytic charging plateaus observed at ∼2.05 V (extraction process)
reduction of the GO. Under strong and prolonged UV represented the Li-deintercalation of LixTiO2.40 The plateaus
illumination in the experiment, most of the oxygen-containing showed the typical electrochemical reaction of anatase TiO2
functional groups of GO are reduced and the TiO2 nano- with lithium ions. The discharge curve showed a sloping
particles remain on the surface of RGO via electrostatic and/or plateau, followed by the TiO2 discharging plateau, which could
van der Waals forces, resulting in the structure as shown in be attributed to the electrochemical behavior of GO/RGO in
Figure 3a. It was reported that the interaction of RGO and the composites.
TiO2 nanoparticles was attributed to the remaining carboxyl Interestingly, Figure 4 shows that the reversible specific
groups at the RGO and its strong adsorption to TiO 2 capacity for TiO2-RGO is ∼310 mAh/g, which is significantly
nanoparticles.28 This is different from our case, in that the higher than the theoretic capacity of TiO2 (168 mAh/g) and
carboxyl groups are mainly distributed on the edge of the GO the theoretic capacity of the composite (200 mAh/g). The
3639 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642
ACS Applied Materials & Interfaces Research Article

theoretic capacity of the as-prepared composite is calculated GO electrodes. The improvement of the TiO2-RGO, in terms
based on fact that the composite contains 5.6% (w/w) RGO of cycling stabilities, may be ascribed to the strong interaction
and 94.4% (w/w) TiO2 and the theoretical capacities are 168 between the TiO2 nanoparticles and RGO sheets.
and 744 mAh/g for TiO2 and graphene, respectively. This Figure 6 shows the rate capabilities of TiO2, TiO2-GO, and
suggests that RGO could also contribute the specific capacity to TiO2-RGO at various current densities between 100 mA/g and
the LIBs. This can be explained by the previous findings that
the Li ion can be stored reversibly in the defects of the edge site
and internal defects on RGO.41 In the case of pure graphene,
the Li-ion storage capacity in graphene could decrease rapidly,
because of the restacking of the graphene nanosheets. In strong
contrast, the restacking of the RGO sheets are effectively
prevented by the anchored TiO2 particles as shown in the TEM
image in Figure 3. Consequently, the reversibly capacity
provided by the RGO are maintained.
In addition, the first discharge and charge capacities are,
respectively, 395 and 160 mAh/g for the TiO2, 536 and 220
mAh/g for the TiO2-GO, and 546 and 296 mAh/g for the
TiO2-RGO. The TiO2-RGO shows a higher Coulombic
efficiency (54%) in the first cycle. In the case of the TiO2
and TiO2-GO, in contrast, their Coulombic efficiency is lower Figure 6. Rate capabilities of TiO2, TiO2-GO, and TiO2-RGO at
(both ∼40%), corresponding to a high irreversible capacity. various current densities.
The large initial capacity loss mainly results from the
decomposition of electrolyte in the low-potential region and
forming a passivation layer (solid electrolyte interface, SEI) on 1600 mA/g. In particular, the reversible capacity of the
the surface of active materials. Nanostructured electrode composite sample is ∼215 mAh/g at a current density of 200
materials normally possess better battery performance, in mA/g. The cycling performance, in terms of the specific
terms of cycle life, rate capability, specific capacity, and energy capacity, gradually declines with the increase of the current
density, because of the much larger surface area. However, the densities for all the samples. For example, at current densities of
large surface could lead to significant side reaction (i.e., 800 and 1600 mA/g, the discharge capacity of the composite
decomposition of the electrolyte). In our case, the formation of sample reduces to 160 and 100 mAh/g, respectively. In all
the composite can effectively restrain the side reaction, which cases, the TiO2-RGO composite samples exhibits better rate
can be demonstrated by the fact that the Coulombic efficient of capability, compared with the TiO2 and TiO2-GO samples.
the composites sample was higher than that of the TiO2 and Moreover, when the current density is set back to the initial 100
TiO2-GO in the initial cycle. mA/g, the specific capacity of the TiO2-RGO samples is fully
We compare the electrochemical properties of the TiO2- recovered. This is beneficial from the strong bonding between
RGO composite, pure TiO2, and TiO2-GO, using 2032 coin- the TiO2 nanoparticles and the RGO and rate capability
type cells to highlight the superiority of the TiO2-RGO contribution from RGO.
composite as anode material in LIBs. The cycling stabilities of Table 1 shows the comparison of the electrochemical
the TiO2-RGO nanocomposite, pure TiO2, and TiO2-GO are performance of the LIBs based on TiO2-RGO or TiO2-
shown in Figure 5. The TiO2-RGO sample shows great graphene composite composites in the literature. The as-
prepared TiO2-RGO is superior to some reported TiO2/
reduced graphene oxide or graphene, in terms of reversible
capacity, and comparable with others, in terms of the rate
capability in the literature.42,43 The excellent electrochemical
performance suggests that the photocatalytic reduction method
is a simple, efficient approach for the preparation of TiO2-RGO
composite.
Lithium-ion insertion/extraction properties of the as-
prepared electrodes were investigated using the CV technique
(see Figure 7). In the first three cycles, all of the samples exhibit
a pair of cathodic/anodic peaks at ∼1.60 and 2.05 V, which can
be ascribed to the Li-ion insertion/extraction in an anatase
TiO2 lattice, respectively. The cathodic/anodic peaks are in
Figure 5. Specific capacities of the TiO2, TiO2-GO, and TiO2-RGO at accordance with the discharge/charge profiles in Figure 4. The
a charge−discharge rate of 100 mA/g. upholding of the cathodic/anodic peak positions suggests that
the reversibility of the TiO2 was well-maintained in the TiO2-
enhancement of the capacity retention in comparison with the RGO composite. For the TiO2-GO and TiO2-RGO samples,
pure TiO2 and TiO2-GO samples. The reversible discharge and we can observe that a prominent peak is located at ∼0.6 V in
charge capacities are, respectively, 310 and 270 mAh/g for the first cycle, because of the formation of SEI layer at the
TiO2-RGO, 220 and 161 mAh/g for TiO2-GO, and 150 and surface of the anodes. These peaks disappear during the
128 mAh/g for the TiO2. Even up to 100 cycles, the TiO2-RGO subsequent cycle, because the SEI layer can isolate the anodes
sample still preserves better electrochemical lithium storage from the electrolyte, thereby preventing further decomposition
performance (more than 200 mAh/g) than the TiO2 and TiO2- of the electrolyte.
3640 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642
ACS Applied Materials & Interfaces Research Article

Table 1. Comparison of the Electrochemical Performance of the As-Prepared TiO2-RGO with Products Reported in the
Literature
material preparing method reversible capacity (mAh/g) rate capability (mAh/g) voltage window (V) reference
anatase/RGO photocatalytic reduction 310 100 @1600 mA/g 0.01−3.0 this work
anatase/RGO hydrothermal 295.9 97.7 @2000 mA/g 0−3.0 43
graphene/TiO2 thermal reduction ∼170 110 @10C 1.0−3.0 42

curves show similar characteristics, including a depressed

semicircle in high and middle frequency range and an inclined
line in the low frequency range, corresponding respectively to
the charge-transfer resistance associated with the surface
properties of insertion materials and the solid-state diffusion
resistance of lithium ions within the host.44 The ohmic
resistances are calculated as 13, 8.5, 3.5 and the charge-transfer
resistance are 280, 179, and 97 for TiO2, TiO2-GO, and TiO2-
RGO, respectively. These results demonstrate that the
conductivity of the graphene sheets was recovered by the
TiO2 photocatalytic reduction of the oxygen-containing
functional group on the GO sheets. Overall, the utilization of
the RGO sheets facilitates the charge-transfer process in the
LIBs.

4. CONCLUSIONS
The insufficient conductivity and theoretical capacity of TiO2
are the main drawbacks for practical application in Li-ion
batteries. In this work, the TiO2-reduced graphene oxide
(TiO2-RGO) nanocomposite was photocatalytically synthe-
sized. The battery performance was significantly improved
because of the unique combination of the TiO2 nanoparticles
and the RGO sheets. On one hand, the conductivity of the
TiO2-RGO nanocomposite was enhanced by anchoring the
TiO2 on the RGO sheets and removing the oxygen-containing
functional groups, compared with that of corresponding the
pure TiO2 and TiO2-GO samples. On the other hand, the
TiO2-RGO sample exhibited much higher specific capacity than
Figure 7. Cyclic voltammograms (CVs) of the TiO2, TiO2-GO, and the pure TiO2 sample, because the RGO could also contribute
TiO2-RGO at a scan rate of 0.1 mV/s. specific capacity without the problem of graphene sheet
restacking. This work suggests that the TiO2-based photo-
catalytic method could be a simple, low-cost, and efficient
AC impedance measurements are performed for all the
approach for the large-scale production of anode materials for
samples at the charged state and the corresponding Nyquist
Li-ion batteries.

■
plots are shown in Figure 8. The inset is the equivalent circuit,
in which RΩ is the total resistance of the electrolyte, separator, AUTHOR INFORMATION
and electrical contacts; Rct is the charge-transfer resistance; and
Cdl represents the double-layer resistance. In addition, Zw is the Corresponding Author
Warburg impedance, reflecting the diffusion of Li ions. All the *Tel.: 61-7-5552 8155. Fax: 61-7-5552 8067. E-mail: s.zhang@
griffith.edu.au.
Notes
The authors declare no competing financial interest.

■ ACKNOWLEDGMENTS
The authors acknowledge the financial support of the ARC
Discovery Grants from the Australian Research Council.

■ REFERENCES
(1) Chen, J. S.; Archer, L. A.; Lou, X. W. J. Mater. Chem. 2011, 21,
9912−9924.
(2) Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D.
Energy Environ. Sci. 2011, 4, 3243−3262.
(3) Lafont, U.; Carta, D.; Mountjoy, G.; Chadwick, A. V.; Kelder, E.
M. J. Phys. Chem. C 2009, 114, 1372−1378.
Figure 8. Nyquist plots of the TiO2, TiO2-GO, and TiO2-RGO (4) Wang, J.; Zhou, Y.; Hu, Y.; O’Hayre, R.; Shao, Z. J. Phys. Chem. C
samples. 2011, 115, 2529−2536.

3641 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642

ACS Applied Materials & Interfaces Research Article

(5) Xiao, X. C.; Liu, P.; Wang, J. S.; Verbrugge, M. W.; Balogh, M. P. (37) Falaras, P.; Hugot-Le Goff, A.; Bernard, M. C.; Xagas, A. Sol.
Electrochem. Commun. 2011, 13, 209−212. Energy Mater. Sol. Cells 2000, 64, 167−184.
(6) Yoo, E.; Kim, J.; Hosono, E.; Zhou, H.-s.; Kudo, T.; Honma, I. (38) Zhang, W. L.; Liu, Y. D.; Choi, H. J.; Kim, S. G. ACS Appl.
Nano Lett. 2008, 8, 2277−2282. Mater. Interfaces 2012, 4, 2267−2272.
(7) Guo, P.; Song, H. H.; Chen, X. H. Electrochem. Commun. 2009, (39) Lindstrom, H.; Södergren, S.; Solbrand, A.; Rensmo, H.; Hjelm,
11, 1320−1324. J.; Hagfeldt, A.; Lindquist, S.-E. J. Phys. Chem. B 1997, 101, 7717−
(8) Ding, S.; Chen, J. S.; Luan, D.; Boey, F. Y. C.; Madhavi, S.; Lou, 7722.
X. W. Chem. Commun. 2011, 47, 5780. (40) van de Krol, R.; Goossens, A.; Meulenkamp, E. A. J. Electrochem.
(9) Shen, L.; Zhang, X.; Li, H.; Yuan, C.; Cao, G. J. Phys. Chem. Lett. Soc. 1999, 146, 3150−3154.
2011, 2, 3096−3101. (41) Zhu, X.; Zhu, Y.; Murali, S.; Stoller, M. D.; Ruoff, R. S. ACS
(10) Liu, D.; Zhang, Y.; Xiao, P.; Garcia, B. B.; Zhang, Q.; Zhou, X.; Nano 2011, 5, 3333−3338.
Jeong, Y.-H.; Cao, G. Electrochim. Acta 2009, 54, 6816−6820. (42) Ding, Y.-H.; Zhang, P.; Ren, H.-M.; Zhuo, Q.; Yang, Z.-M.;
(11) Das, S. K.; Bhattacharyya, A. J. J. Phys. Chem. C 2009, 113, Jiang, Y. Mater. Res. Bull. 2011, 46, 2403−2407.
17367−17371. (43) Cao, H.; Li, B.; Zhang, J.; Lian, F.; Kong, X.; Qu, M. J. Mater.
(12) Armstrong, G.; Armstrong, A. R.; Bruce, P. G.; Reale, P.; Chem. 2012, 22, 9759−9766.
Scrosati, B. Adv. Mater. 2006, 18, 2597−2600. (44) Bueno, P. R.; Leite, E. R. J. Phys. Chem. B 2003, 107, 8868−
(13) Wang, Y. F.; Wu, M. Y.; Zhang, W. F. Electrochim. Acta 2008, 8877.
53, 7863−7868.
(14) Chen, J. S.; Tan, Y. L.; Li, C. M.; Cheah, Y. L.; Luan, D.;
Madhavi, S.; Boey, F. Y. C.; Archer, L. A.; Lou, X. W. J. Am. Chem. Soc.
2010, 132, 6124−6130.
(15) Wang, D.; Choi, D.; Li, J.; Yang, Z.; Nie, Z.; Kou, R.; Hu, D.;
Wang, C.; Saraf, L. V.; Zhang, J.; Aksay, I. A.; Liu, J. ACS Nano 2009,
3, 907−914.
(16) Fu, L. J.; Liu, H.; Zhang, H. P.; Li, C.; Zhang, T.; Wu, Y. P.; Wu,
H. Q. J. Power Sources 2006, 159, 219−222.
(17) Xu, J. W.; Wang, Y. F.; Li, Z. H.; Zhang, W. J. Power Sources
2008, 175, 903−908.
(18) Zhang, H.; Lv, X. J.; Li, Y. M.; Wang, Y.; Li, J. H. ACS Nano
2010, 4, 380−386.
(19) Li, Y.; Lv, X.; Lu, J.; Li, J. J. Phys. Chem. C 2010, 114, 21770−
21774.
(20) Paek, S.-M.; Yoo, E.; Honma, I. Nano Lett. 2008, 9, 72−75.
(21) Wu, Z.-S.; Ren, W.; Wen, L.; Gao, L.; Zhao, J.; Chen, Z.; Zhou,
G.; Li, F.; Cheng, H.-M. ACS Nano 2010, 4, 3187−3194.
(22) Li, B.; Cao, H.; Shao, J.; Qu, M.; Warner, J. H. J. Mater. Chem.
2011, 21, 5069−5075.
(23) Hummers, W. S.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80,
1339−1339.
(24) Marcano, D. C.; Kosynkin, D. V.; Berlin, J. M.; Sinitskii, A.; Sun,
Z.; Slesarev, A.; Alemany, L. B.; Lu, W.; Tour, J. M. ACS Nano 2010, 4,
4806−4814.
(25) Williams, G.; Seger, B.; Kamat, P. V. ACS Nano 2008, 2, 1487−
1491.
(26) Gao, X.; Jang, J.; Nagase, S. J. Phys. Chem. C 2009, 114, 832−
842.
(27) Vinodgopal, K.; Neppolian, B.; Lightcap, I. V.; Grieser, F.;
Ashokkumar, M.; Kamat, P. V. J. Phys. Chem. Lett. 2010, 1, 1987−
1993.
(28) Bell, N. J.; Yun, H. N.; Du, A. J.; Coster, H.; Smith, S. C.; Amal,
R. J. Phys. Chem. C 2011, 115, 6004−6009.
(29) Jiang, D.; Zhao, H.; Zhang, S.; John, R. J. Phys. Chem. B 2003,
107, 12774−12780.
(30) Lightcap, I. V.; Kosel, T. H.; Kamat, P. V. Nano Lett. 2010, 10,
577−583.
(31) Meyer, T. J.; Meyer, G. J.; Pfennig, B. W.; Schoonover, J. R.;
Timpson, C. J.; Wall, J. F.; Kobusch, C.; Chen, X. H.; Peek, B. M.
Inorg. Chem. 1994, 33, 3952−3964.
(32) McAllister, M. J.; Li, J.-L.; Adamson, D. H.; Schniepp, H. C.;
Abdala, A. A.; Liu, J.; Herrera-Alonso, M.; Milius, D. L.; Car, R.;
Prud’homme, R. K.; Aksay, I. A. Chem. Mater. 2007, 19, 4396−4404.
(33) Moon, I. K.; Lee, J.; Ruoff, R. S.; Lee, H. Nat. Commun. 2010, 1,
73.
(34) Fan, Y.; Lu, H. T.; Liu, J. H.; Yang, C. P.; Jing, Q. S.; Zhang, Y.
X.; Yang, X. K.; Huang, K. J. Colloids Surf., B 2011, 83, 78−82.
(35) Zhang, H.; Fan, X.; Quan, X.; Chen, S.; Yu, H. Environ. Sci.
Technol. 2011, 45, 5731−5736.
(36) Li, N.; Liu, G.; Zhen, C.; Li, F.; Zhang, L.; Cheng, H.-M. Adv.
Funct. Mater. 2011, 21, 1717−1722.

3642 dx.doi.org/10.1021/am300722d | ACS Appl. Mater. Interfaces 2012, 4, 3636−3642