[Link].

de
[Link]

Structure, Synthesis, and Applications of TiO2 Nanobelts

REVIEW
Zhenhuan Zhao, Jian Tian, Yuanhua Sang,* Andreu Cabot,* and Hong Liu*

nanomaterial-based technologies having

TiO2 semiconductor nanobelts have unique structural and functional proper- been commercialized.[1] In particular,
ties, which lead to great potential in many fields, including photovoltaics, photo- since TiO2 was found to be a chemically
catalysis, energy storage, gas sensors, biosensors, and even biomaterials. A stable, low cost, environment-friendly,
review of synthetic methods, properties, surface modification, and applications and efficient hydrogen production photo-
catalyst in 1972,[2] it has been widely used
of TiO2 nanobelts is presented here. The structural features and basic proper-
for several applications, such as photo-
ties of TiO2 nanobelts are systematically discussed, with the many applications catalysis,[3] solar cells,[4] gas sensors,[5] and
of TiO2 nanobelts in the fields of photocatalysis, solar cells, gas sensors, biosen- biosensors.[6] Most initial investigations
sors, and lithium-ion batteries then introduced. Research efforts that aim to focused on granulated TiO2 and dem-
overcome the intrinsic drawbacks of TiO2 nanobelts are also highlighted. These onstrated the excellent performance of
efforts are focused on the rational design and modification of TiO2 nanobelts by TiO2 nanoparticles in removing organic
pollutants and generating hydrogen by
doping with heteroatoms and/or forming surface heterostructures, to improve
water splitting, as well as selective col-
their desirable properties. Subsequently, the various types of surface hetero- lection or charge transport in solar cells,
structures obtained by coupling TiO2 nanobelts with metal and metal oxide gas sensors and biosensors.[7] However,
nanoparticles, chalcogenides, and conducting polymers are described. Further, granulated TiO2 nanoparticles have also
the charge separation and electron transfer at the interfaces of these hetero- shown some unavoidable disadvantages.
For photocatalysis or solar cell applica-
structures are also discussed. These properties are related to improved sensi-
tion, the large interface area associated
tivity and selectivity for specific gases and biomolecules, as well as enhanced with small particles strongly affects charge
UV and visible light photocatalytic properties. The progress in developments of transport properties. Besides, the high
near-infrared-active photocatalysts based on TiO2 nanobelts is also highlighted. concentration of surface defects associated
Finally, an outline of important directions of future research into the synthesis, with nanoparticles results in significant
modification, and applications of this unique material is given. charge recombination.[8] Another serious
disadvantage of TiO2 nanoparticles for
practical photocatalytic applications is the
1. Introduction high recycling cost when used in aqueous
media. These drawbacks limit further improvement of the per-
The increasing research activities and great progresses in formance of TiO2 nanoparticles in several applications.
fundamental aspects and practical applications of nano- 1D nanostructures have shown distinctive advantages with
materials and nanofabrication technologies are changing the respect to 0D and 2D materials.[9] 1D nanostructures have
world. The amazing properties of nanomaterials have revolu- the smallest dimensionality that allows two key properties for
tionized many industries, with numerous nanomaterials and photocatalysis and nanodevices to be combined in the same
material: fast electron transport and effective charge transfer.[10]
Thus, 1D nanostructures are both an ideal model system for
Dr. Z. Zhao, Dr. J. Tian, Dr. Y. Sang, Prof. H. Liu exploring dimensionality dependent electronic, optoelectronic
State Key Laboratory of Crystal Materials and electromechanical properties, and an ideal building block
Shandong University for the production of high performance multifunctional het-
Jinan 250100, P. R. China
E-mail: sangyh@[Link]; hongliu@[Link] erostructures. In particular, since the first synthesis of belt-
Dr. Z. Zhao, Prof. H. Liu like ZnO nanostructures,[11] 1D nanobelts have attracted much
Beijing Institute of Nanoenergy and Nanosystems attention due to their unique structural and functional proper-
Chinese Academy of Science ties. Among the different materials obtained with a belt-like
Beijing 100083, P. R. China geometry, TiO2 nanobelts — with a large concentration of
Prof. A. Cabot exposed active facets and exceptional electronic, optical and
Catalonia Institute for Energy Research – IREC
Sant Adria del Besos
mechanical properties — have raised particular interest and
Barcelona 08930, Spain have been proposed for a remarkably wide range of applica-
E-mail: acabot@[Link] tions.[12] Nowadays, thanks to the large effort in developing
Prof. A. Cabot synthesis methods to produce this material, single crystal TiO2
Institució Catalana de Recerca i Estudis Avançats – ICREA nanobelts with exceptional crystallinity and a very low concen-
Barcelona 08010, Spain tration of defects and dislocations are available. These nano-
DOI: 10.1002/adma.201405589 structures are characterized by exceptional charge transport

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 1
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

Figure 1. a) Typical scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HRTEM) images of rutile TiO2 nano-
belts obtained by acid-hydrothermal process. Reproduced with permission.[16] Copyright 2011, RSC. b) Typical transmission electron microscopy (TEM)
and SEM images of anatase TiO2 nanobelts. The inset energy dispersive X-ray (EDS) spectrum reveals the presence of only Ti and O in the nanobelt
(Si comes from the substrate). The inset high-magnification SEM image shows the width and fractured cross-section of a nanobelt. Reproduced with
permission.[17]

properties, with fast charge transportation along the axial direc- synthetic methods, and applications. The wide applications of
tion, and an easier recyclability than granule-like nanoparticles. TiO2 nanobelt are highly related to the properties of TiO2 nano-
Furthermore, their electronic performance can be further con- belt. In this Review, the basic properties of TiO2 nanobelts are
trolled and enhanced through quantum confinement effects. At introduced before the synthetic methods, as we want readers
the same time, chemically stable and ultralong TiO2 nanobelts firstly to fully understand the properties. 2) To clearly address
can be manipulated for tuning of functional properties and the design principles of TiO2-nanobelt-based heterostructures
optimization for numerous potential applications, such as con- and band-engineered nanobelts, in order to aid rational design
tinuous photocatalysis and sensors. and construction of specific nanostructures using proper mate-
Nevertheless, although TiO2 nanobelts have great poten- rials. 3) To highlight the very recent progress in the develop-
tial as photocatalysts and solar cell electrodes, they also pos- ment of full solar light photocatalysis based on TiO2 nanobelts.
sess some intrinsic drawbacks. One main concern is the wide
optical bandgap of TiO2, which limits its light harvesting capa-
bility to the UV light range, leaving about 95% of the solar 2. Structure and Basic Properties
light energy wasted.[13] A second major limitation is the low
of TiO2 Nanobelts
photo-electronic coupling ability characterizing the poor photo-
active facets of TiO2 nanobelts. To overcome these drawbacks, TiO2 has four crystalline polymorphs, including rutile, anatase,
and be able to develop high performance photo-electric con- brookite, and TiO2 (B). Rutile- and anatase-phase TiO2 nano-
version materials and devices based on TiO2 nanobelts, is a belts have been most intensively investigated for photocatalysis
present major challenge. In recent years, many works aiming applications because of their better catalytic activity and easier
to overcome the optoelectronic limitations of TiO2 nanobelts synthesis. Therefore, in this section, we focus mainly on the
have been reported. These works are based on several design structural characteristics and properties of rutile and anatase
principles, such as enlarging the photocatalytic active surface, TiO2 nanobelts.
enhancing/broadening the light harvesting range, forming
Schottky or p–n junctions, or engineering their band structure
to match particular energy levels. The very large surface area 2.1. Structural Characteristics
of TiO2 nanobelts also provides a great platform to assemble a
second nanomaterial to improve the relatively poor photo-elec- Rutile and anatase TiO2 nanobelts have an octahedral arrange-
tric conversion ability of bare TiO2 nanobelts. ment (very similar to that of TiO6) with Ti and O atoms
The goal of this article is to provide a complete picture of the having the same coordination in their unit cell.[14] The layered
state of the art in TiO2 nanobelts by reviewing the most impor- arrangements of octahedrons in these two phases facilitates
tant and most recent papers. The main areas of focus of this their growth in a belt-like geometry.[15] Both rutile and anatase
review article are as follows. 1) To provide a full introduction to nanobelts display a similar morphology, i.e., rectangular cross
TiO2 nanobelts including their basic properties, commonly used section with six exposed smooth facets (Figure 1). Rutile TiO2

2 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
nanobelts have very small thicknesses of about 20 nm, widths The band structure of TiO2 nanobelts is not only affected by
of 50–100 nm and can be several micrometers in length.[16] their diameter, but also by their surface physical and chemical
Rutile nanobelts, having a tetragonal phase with lattice properties, and their growth orientation.[22,24] Different growth
parameters a = b = 0.459 nm, and c = 0.296 nm are obtained orientations result in different facet terminations, i.e, termi-
with very high crystalline quality. Although the growth direction nated with Ti or O atoms. However, it is found that both arm-
for 1D nanostructures depends on the growth method used, for chair and zigzag TiO2 nanobelts with bandgaps larger than
rutile TiO2 nanobelts, the growth direction is generally the [001] 2.93 eV prefer stable, oxygen-terminated facets.
direction. The geometric parameters of anatase TiO2 nano- TiO2 nanobelts without oxygen vacancies are proven to be
belts are very close to those of rutile TiO2 nanobelts, having spin-unpolarized wide-bandgap semiconductors with variable
widths of about 50–200 nm and thicknesses of 20–50 nm.[17] bandgaps depending upon the growth direction and width,
The crystalline structure of anatase TiO2 nanobelts can also as shown in Figure 2b. On the other hand, TiO2 nanobelts
be assigned to a tetragonal phase, but with different lattice containing oxygen vacancies (Vo) show narrower bandgaps,
parameters: a = b = 0.3766 nm and c = 0.9486 nm. The growth depending on the concentration of oxygen vacancies. Thus, the
direction of anatase TiO2 nanobelts is different from that of wide bandgap of TiO2 nanobelts can be narrowed by intention-
rutile TiO2 nanobelts. After recording a series of selected area ally introducing oxygen vacancies.
electron diffraction (SAED) patterns of single anatase TiO2 It is well known that anatase TiO2 is more photoactive than
nanobelts in different crystallographic orientations, keeping the rutile TiO2. Because of this, details of the electronic structure of
[101] axis horizontal,[18] it was found that the growth direction rutile TiO2 nanobelts are hardly reported. Taking into account
of anatase TiO2 nanobelts was the [010] direction, with a typical the important roles that electron mobility and lifetime play in
spacing of 0.38 nm. photocatalysis, the main reason for this difference in photo-
activity is probably the 40-times-larger electron mobility of
anatase TiO2 compared with rutile TiO2.[25] In this [010] direc-
2.2. Basic Properties of TiO2 Nanobelts tion, anatase TiO2 nanobelts are characterized by electrical
conductivities of about 10−7 Ω−1 cm−1 (Figure 2c),[17] well above
While the differences between the crystalline structure of those of anatase TiO2 nanoparticles, at 10−9 Ω−1 cm−1.[18]
rutile and anatase TiO2 nanobelts are tiny, they are sufficient Electron lifetimes depend on the kinetics of trapping and
to result in dissimilar electronic structure and optical, mechan- detrapping of electrons at states located in the bandgap of the
ical and surface properties. These properties affect the perfor- oxide, and significantly influence the electron transport and
mance of rutile and anatase TiO2 nanobelts in their multiple electron diffusion.[26] TiO2 nanobelts have significantly higher
applications. diffusion coefficients and much lower charge recombination
than granulated TiO2 (Figure 2d).[27] Because of the longer
electron lifetime, larger electron diffusion length, and lower
2.2.1. Electronic Band Structure charge carrier recombination, TiO2 nanobelts have significantly
reduced charge loss during transportation, improving charge
Electronic properties are known to control light absorbance, transfer and charge collection efficiency with respect to granu-
redox potential, and charge carrier mobility, and consequently lated TiO2. Nanobelts also have almost two-fold higher effective
to strongly affect the functional properties of semiconduc- donor concentrations than nanoparticles, with concentration of
tors.[19] Full understanding of the electronic band structures 3.85 × 1019 cm−3, compared with 1.90 × 1019 cm−3.[27]
of 1D TiO2 nanostructures, especially those of TiO2 nanobelts,
is critical for their application in several fields, such as photo-
catalysis, solar energy conversion or gas sensing. The wide 2.2.2. Optical Properties
bandgaps of bulk anatase TiO2 (3.2 eV) and bulk rutile TiO2
(3.0 eV) are related to the existence of the O 2p and Ti 3d states The light absorption ability of a semiconductor is mainly related
in the valence band (VB) and the conduction band (CB).[20] 1D to interband electron transitions and is thus controlled by the
TiO2 nanostructures usually display thickness-dependent band- bandgap. Excitation of an electron to a point within the conduc-
gaps.[21] When the thickness of the TiO2 nanobelt is less than tion band occurs on absorbtion of a photon with energy equal
2.5 nm, the bandgap is enlarged to a value above that of bulk to the sum of the bandgap energy and the energy within the
anatase and bulk rutile TiO2, because of quantum confinement conduction band, minus the energy of the hole left within the
effects arising with the decrease in thickness of the nanobelt valence band.[28] These electron-hole pairs are those required
(Figure 2a).[22] However, in general, thicker TiO2 nanobelts are for a photoreaction in a photocatalyst, or those extracted in a
produced, and are therefore usually characterized by bandgaps photovoltaic device.[29] For nanometer-sized semiconductors,
narrower than the 3.4 eV theoretically predicted for the 2.5 nm the optical bandgap may be affected by the size and surface
TiO2 nanobelts. The bandgaps of TiO2 nanobelts are typically states of the nanoparticles. TiO2 shows an excellent UV light
much lower than those of TiO2 nanotubes, which are estimated photocatalytic activity;[30] however, its wide bandgap permits
to be about 3.87 eV for room-temperature aqueous colloids of TiO2 to only absorb light with wavelength below about 400 nm,
TiO2 nanotubes. The wider bandgaps of TiO2 nanotubes are which only accounts for 4% of the solar energy reaching the
very close to the bandgaps of 2D titanate nanosheets, causing earth surface.
the 1D TiO2 nanotube to display apparent 2D behavior at room Absorbance significantly depends on the TiO2 crystal phase
temperature.[23] and therefore on the calcination temperature used during the

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 3
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

Figure 2. a) Energy gap evolution of ZTiO2NRs (zigzag TiO2 nanoribbons, triangles) and ATiO2NRs (armchair TiO2 nanoribbons, squares) as a function
of nanobelt or ribbon width. The solid symbols represent the nanobelts with N = 2m while the hollow symbols represent the nanobelts with N = 2m
+ 1, where N = Nz or NA and m is a positive integer. The bandgaps are fitted to the formula Eg = a + b/W, where the fitting parameters a and b are
3.49 and 0.82 eV Å for ATiO2NRs with even NA; 3.49 and –2.07 eV Å for ATiO2NRs with odd NA; 3.62 and 1.23 eV Å for ZTiO2NRs with even Nz; and
3.62 and –1.68 eV Å for ZTiO2NRs with odd Nz. Reproduced with permission.[22] Copyright 2010, ACS. b) Isosurfaces of spin density ρ = ρ↑ – ρ↓ (top)
and band structures (bottom) of 6-ZTiO2−x NRs with Vo-defected one edge (left), two edges (middle), and hydrogenated Vo defects at two edges (right).
The isovalue is 0.01 |e|/Å3. The yellow and cyan surfaces represent spin-up and spin-down, respectively. The band lines of spin-up and spin-down
branches are represented by the red and blue lines. The energy at the Fermi level (EF) is set to zero. Reproduced with permission.[22] Copyright 2010,
ACS. c) Current–voltage (I–V) characteristics of a TiO2-nanobelt-based membrane showing a constant resistivity. Reproduced with permission.[17]
d) Electron lifetimes in TiO2 nanoparticles (TNP), TiO2 nanobelts (TNR) and a mixture of TiO2 nanoparticles and TiO2 nanobelts (TPR10). Reproduced
with permission.[27] Copyright 2011, ACS. e) Diffusion coefficient and diffusion length for electrons in TiO2 nanoparticle (TNP), TiO2 nanobelt (TNR),
and TPR10. Reproduced with permission.[27] Copyright 2011, ACS.

synthesis process. It has been reported by Zhou et al. that the than anatase TiO2, more investigation is needed for better under-
light absorption edge of TiO2 nanobelts changes with the calcina- standing of the different optical and optoelectronic properties
tion temperature, while the crystalline phase evolves from as-syn- of rutile TiO2 and anatase TiO2 nanobelts. The light absorption
thesized H2Ti3O7, to TiO2(B), to anatase, and finally to rutile.[31] of TiO2 nanotubes is different from that of TiO2 nanobelts. The
This dependence of the bandgap on the annealing temperature wider bandgap of the TiO2 nanotubes makes the position of the
up to 800 ºC has been also demonstrated by Fu et al.[32] The cut-off edge of light absorption of TiO2 nanotubes blue-shifted
bandgap of TiO2 nanobelts obtained at 100, 400, 600, 800, 900 with respect to that of the TiO2 nanobelts. At the same time, the
and 1000 °C are 2.85, 2.8, 2.9, 3.05, 3.1 and 2.9 eV, respectively, absorption characteristics of TiO2 nanotubes show apparent 2D
slightly smaller than that of bulk TiO2. It is suggested that the behavior, very similar to that of TiO2 nanosheets.
increase in the bandgap values is associated with the phase tran- In order to build a high performance photocatalyst, sup-
sition from TiO2(B), with a bandgap value of 2.8 eV, to anatase, pressing the recombination of charge carriers — occurring
with a bandgap of 3.1 eV. However, notice that materials with both at the surface and in the bulk of the semiconductor — is
a combination of phases are usually obtained. As an example, critical. Both anatase and rutile TiO2 nanobelts have intensive
the sample calcined at 1000 °C contains both anatase and rutile light emission,[16,33] indicating relatively high radiative recom-
phases.[31] Notice also that the rutile nanobelts obtained after bination efficiencies of the photogenerated charge carriers.[34]
the highest temperature process have higher degree of crystal- The presence of dislocations and local defects acting as electron
linity than the other TiO2 phases. Direct measurements of the and hole trapping centers is responsible for both a deterioration
light absorption ability of rutile TiO2 nanobelt have been hardly of the material electrical properties and an increase in the non-
reported. Though rutile TiO2 has lower photocatalytic activity radiative recombination of the charge carriers.[18]

4 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
and width of 210 nm was suggested to be
11.87 GPa,[42] which is considerably lower
than that of bulk TiO2 (280 GPa). The rea-
sons behind the large difference between the
rutile and anatase phases is attributed to the
different levels of vacancies within the nano-
belts obtained from different preparation
methods.[43] In this [010] direction, the con-
centration of the free surface atoms is sug-
gested to determine the mechanical proper-
ties of TiO2 nanobelts.

2.2.4. Surface Properties

The photocatalytic activity and selectivity

strongly depend on surface properties. As
an example, the chemical dissociation of
water molecules is energetically favored on
the (001) plane; in contrast, water molecules
Figure 3. a) Schematic diagram of the test of a suspended TiO2 nanobelt-beam by an atomic are preferentially physically adsorbed on the
[44]
force microscope (AFM) with a nanoindenter. b) Fload varying with time. c) SEM images of (101) plane of TiO2. The dissociated water
suspended TiO2 nanobelt-beams. d) Deflection along the suspended nanobelts-beam at three molecule probably facilitates the transfer of
fixed load forces. a–d) Reproduced with permission.[38] Copyright 2012, RSC. photoinduced charge carriers and promotes
the formation of reactive radicals. It has been
2.2.3. Mechanical Properties suggested that this is the reason for the much more effective
photoreaction occurring at the (001) plane, compared with that
One important application of TiO2 nanobelts is their use as at the (101) plane.[45]
building blocks for fabrication of various micro/nano devices The surface energy of the different facets is one main
such as various single belt-based sensors or electronics.[35] parameter determining its photoactivity. The surface energy
The advantages of TiO2 nanobelts in this context are ascribed of rutile TiO2 facets is γ(001) > γ(100) > γ(101) (0.90 J m−2 >
not only to their unique geometry and electronic and optical 0.53 J m−2 > 0.44 J m−2). Consistently with this order, Liu et al.
properties, but also to their excellent mechanical properties,[36] have reported an enhanced photoactivity of TiO2 with exposed
which are totally different from the bulk material.[37] Compared (001) and (110) facets.[46] For the photodegradation of organic
with TiO2 nanobelts, TiO2 nanotubes are very fragile and are pollutants, it is also well accepted that the high-energy (001)
easily broken up by mechanical forces. facet is more effective than the (101) facet in the dissociative
Present studies on the mechanical properties of TiO2 mainly adsorption of reactant molecules. This is attributed to the
focus on rutile TiO2 nanobelts, and few works have reported reduction of organic pollutant activation energy on the (001)
the mechanical behavior of anatase TiO2 nanobelts. This may facet. The chemically dissociated molecules react more effi-
be related to the synthetic methods that will be discussed in ciently with photoproduced hydroxyl radicals on the surface
the following sections. Wang and co-workers have investigated of TiO2 and in consequence can be rapidly converted to inac-
the mechanical properties of TiO2 nanobelts with thicknesses tive surface hydroxyls. In other words, the photodegradation
of 60 nm, lengths of 8 µm and widths of 1.6 µm.[38] They kinetics and the charge transfer process can be enhanced by
found that a suspended nanobelt breaks under a critical frac- the dissociated adsorption of pollutant molecules. However,
ture stress of about 4.3 µN (Figure 3). The fracture strength TiO2 nanostructures with exposed energetic (001) facets are
of the TiO2 nanobelt was reported to be about 250 GPa, and hard to produce, and most synthetic routes reported involve
the Young’s modulus about 120 GPa. The Young’s modulus the addition of highly toxic chemicals to stabilize the active
of TiO2 nanobelts with an average cross-section dimension (001) facet.
of about 30 nm was about 260 GPa. For thicker TiO2 nano- As expected, rutile and anatase TiO2 nanobelts have different
belts, the bending modulus rapidly decreases with the increase catalytic properties, in a large part due to the different exposed
of the thickness of the nanobelts.[39] As a result, the cor- crystalline facets which rule the photocatalytic activity.[45] As
responding shear modulus is suggested to be in the range an example, superoxide radicals act as an intermediate in the
of 0.07–0.4 GPa, ruled by the interlayer interactions and the oxidization of organic molecules.[47] The superoxide radical is
defect density.[40] reported to originate from the adsorption of oxygen molecules
The Young’s modulus of rutile TiO2 nanobelts with an on the surface of TiO2, which subsequently traps the photogen-
average thickness of 30 nm and a half-width of 80 nm was erated electrons. As stated above, anatase TiO2 nanobelts have
measured to be about 360 GPa by nanoindentation, after care- two dominant (101) crystalline facets. Though the (101) facet on
fully subtracting the substrate effect.[41] The average modulus anatase TiO2 is less energetically favored, it has been demon-
of elasticity of anatase nanobelts with a thickness of 294 nm strated to have the ability to promote the adsorption of oxygen

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 5
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

molecules, which is beneficial to the production of superoxide main steps in the formation of TiO2 nanobelts according to the
radicals. following three equations. First, the TiO2 precursor is dissolved
The oxygen chemisorption onto the TiO2 surface not only by breaking the Ti–O–Ti bonds, leading to the formation of
depends on the specific facet, but is also strongly affected by sodium titanate nanobelts. Secondly, the sodium titanate nano-
the surface state of TiO2. It is suggested that the reactivity of the belt is converted to hydrogen titanate nanobelt by ion-exchange.
TiO2(110) facets is primarily governed by the surface species, Finally, the hydrogen titanate nanobelt is annealed to form TiO2
such as oxygen vacancies and bridging hydroxyls,[48] changes in nanobelts.
which may change the surface electronic structure of TiO2.[49]
For annealed TiO2 surfaces, one oxygen vacancy can adsorb two 3TiO2 + 2NaOH → Na 2 Ti 3O7 + H2O (1)
oxygen molecules. When the TiO2 surface is exposed to water,
one oxygen vacancy is converted to two bridging hydroxyls. Na 2 Ti 3O7 + 2HCl → H2 Ti 3O7 + 2NaCl (2)
In addition, different crystalline planes have different reduc-
tion or oxidation potentials, therefore, upon photoexcitation,
photogenerated electrons and holes may migrate to different H2 Ti 3O7 → 3TiO2 + H2O (3)
planes separating from each other.
Proceeding with these three steps, the nanobelt morpholo-
gies of Na2Ti3O7, H2Ti3O7 and TiO2 can be maintained if the
3. Synthetic Methods to Produce TiO2 Nanobelts reaction processes is carefully controlled, as shown in Figure 4a.
TiO2 nanobelts obtained from alkali-hydrothermal process typi-
Nucleation and growth are the two fundamental steps in the cally have an anatase phase with the growth direction along
synthesis of 1D nanomaterials. Based on these two steps, a with the [010] direction, as shown in Figure 4b.[18] It should be
variety of methods for the synthesis of TiO2 nanobelts have pointed out that by carefully controlling the hydrothermal con-
been developed. These methods show different levels of con- ditions, anatase TiO2 nanobelts with a dominant (001) facet can
trol over the synthetic parameters; for example, introducing also be prepared by the alkali-hydrothermal process.[55]
a liquid-solid interface to reduce the symmetry of the seed, The key point for formation of the belt-like morphology is the
employing various templates with 1D morphology, adding cap- formation of sodium titanate nanobelts: thus the mechanism of
ping reagents to control the growth rates of different facets, formation of such sodium titanate nanobelts has been widely
and so on.[10] There are several general approaches to the syn- investigated.[22,57] Figure 5 illustrates the growth mechanism
thesis of TiO2 nanobelts, such as hydrothermal processes,[50] of TiO2 nanobelts. Initially during the hydrothermal process,
electrochemical anodization,[51] and metallorganic chemical six-coordinated monomers [Ti(OH)6]2− are formed and inclined
vapor deposition,[52] which are described in the following to combine with each other form a nucleus. These nuclei con-
sections. sequently grow into nanosheets consisting of layers of octahe-
dral [TiO6] units. Large amounts of Na+ cations are dispersed
among the layers to balance the negative charges of the [TiO6]
3.1. Hydrothermal Methods octahedrals. These layered structures facilitate the replacement
of the Na+ by larger H3O+ cations, weakening the static interac-
As one of the most popular methods toward the synthesis of tion due to the enlarged interlayer distance. As a result, the lay-
TiO2 nanobelts, the hydrothermal method can be an effective ered sodium titanate is gradually exfoliated to form hydrogen
and facile approach to the formation of highly crystallized TiO2 titanate nanosheets that split to form nanobelts, in order to
nanobelts. The hydrothermal process is divided into two routes, release the strong stress and lower the total energy.[28]
i.e., the alkali-hydrothermal process and the acid-hydrothermal The size and morphology of the TiO2 nanobelt can be tuned
process. In the alkali-hydrothermal process, several materials by adding a surfactant. Xie and Shang have reported a solvo-
can be used as the titanium precursors including titanium thermal method using titanium tetraisopropoxide (TTIP) as
oxide powder, titanium butoxide (Ti(OBu)4), titanium chloride a precursor. Ethylenediamine (EDA) and ethylene glycol (EG)
(TiCl4), TiF4, TiS2, and so on.[53] In a typical procedure, the tita- were used to control the morphology.[58] It is found that the
nium precursor is first dispersed into a NaOH aqueous solution morphology of the TiO2 nanobelts was strongly dependent on
with a concentration of 10 M and then hydrothermally heated the relative concentration of EDA. At very low concentrations,
at 180–200 °C for at least 12 h. It is commonly accepted that EDA reduced the diameter but kept the belt form of the TiO2.
sodium titanate nanobelts with similar crystalline structure are In this case, EDA plays a chelating-ligand role, influencing the
formed during the alkali-hydrothermal process. The chemical solubility, reactivity, and diffusivity of the reactants. In this sce-
formula of the sodium titanate nanobelt is determined by the nario, the polarity and coordination ability of solvents like EDA
adopted titanium precursor. For example, TiO2 powder tends can influence the crystal morphology of the final product. On
to form sodium titanate nanobelts in the form of Na2Ti3O7 or further increasing the concentration of EDA, TiO2 is formed
Na2Ti4O9.[17,53b] The Ti(OBu)4 and TiCl4 precursors tend to form with a nanorod morphology, and at very high EDA concentra-
sodium titanate nanobelts in the form of Na2Ti2O4(OH)2, while tions, far exceeding the amount of TTIP, TiO2 assumes a fine
TiF4 is reported to form Na2Ti5O11 nanobelts. These sodium fibril form.
titanate nanobelts have a similar layered structure. It is thought Another, completely different, route for producing TiO2
that the formation process of sodium titanate nanobelts involves nanobelts is the acid-hydrothermal process by the hydrolysis
a dissolution/recrystallization mechanism.[54] There are three of a titanium precursor, e.g., tetrabutyl titanate. The nature of

6 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
Figure 4. a) Typical SEM images of Na2Ti3O7, H2Ti3O7 and TiO2 nanobelts. Reproduced with permission.[17] b) Bright-field TEM image of an anatase
nanobelt and the selected area electron diffraction pattern taken along the [100] direction as well as the HRTEM image. Reproduced with permission.[18]
Copyright 2010, ACS. c) TEM image of a single rutile nanobelt and the selected area diffraction pattern (inset) of a nanobelt prepared using 12 M HCl and
titanium(IV) butoxide by hydrothermal synthesis, as well as the HRTEM image. Reproduced with permission.[56] Copyright 2013, Royal Society of Chemistry.

this route is delicate control of the rate of hydrolysis of titanium the morphological evolution of TiO2 in acid-hydrothermal con-
precursor in acid conditions, usually resulting in the forma- ditions and found that rutile TiO2 nanobelts can be formed
tion of rutile TiO2 nanobelts with the [001] growth direction, as at higher acid concentrations (9–10 M).[16] In a typical proce-
shown in Figure 4c.[56] Zhou et al. systematically investigated dure, 0.2 mL tetrabutyl titanate was homogenously mixed with
9–10 M HCl aqueous solution, and then
hydrothermally treated at 180 °C for 2 h in
a 20 mL Teflon vessel. Compared with the
alkali hydrothermal method, this route has
the advantages of being a one-step synthesis,
with much shorter hydrothermal treatment
time.
The above mentioned dependence of
the morphology of rutile TiO2 nanobelt on
acid concentration is further demonstrated
by Cha et al.[56] In their experiments, a
12 M HCl aqueous solution was used to
prepare rutile TiO2 nanobelts. Interestingly,
they found that the length of the nanobelt
increased with increasing reaction time,
while the nanobelts obtained after longer
times (8 h) were much thinner that those
formed at shorter times (4 h). They observed
the phenomenon of a thin rutile nanobelt
of 20 nm wide splitting into several thin
nanobelts each about 5 nm wide, caused by
a lattice misorientation originating from an
array of dislocations.
A Ti foil has been also used as the tita-
nium source for the hydrothermal synthesis
[59] Before the
Figure 5. Schematic illustration showing the growth mechanism of TiO2 nanobelts. The [TiO6] of TiO2 nanowire or nanobelts.
octahedral consists of one Ti atom and six O atoms, with the Ti atom in the center and six O hydrothermal reaction, the Ti foil was ultra-
atoms in the corners of the octahedral. sonically cleaned using water, acetone, and

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 7
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

ethanol, respectively. It was then placed into the Teflon-lined pre-coated onto the substrate, and the temperature is typically
stainless steel autoclave, which contained 30 mL 1 M aqueous higher than 850 °C. Amin et al. successfully grew TiO2 nano-
NaOH solution. The hydrothermal process was conducted belts at a decreased temperature (850–920 °C) by using nickel
at 220 °C for 24 h in an electric oven. Note that the concen- as the catalyst.[64] In this case, it was found that the formation
tration of the alkali solution required with this Ti precursor of the TiO2 nanostructure was substrate-independent. In this
is much lower than that needed when using TiO2 powder as case, the nanobelt growth is suggested to proceed through a
the precursor. The hydrothermal reaction leads to the forma- vapor–liquid–solid (VSL) growth mechanism. While several
tion of Na2Ti2O4(OH)2 nanobelt arrays, which change into metals like Ni, Au, Ag, Pt, and Pd can assist the growth of TiO2
H2Ti2O4(OH)2 nanobelts by immersion into the HCl solution. nanostructures, Ni and Au are preferred as they result in higher
A post-calcination treatment at 500 °C can easily convert the yields.
H2Ti2O4(OH)2 into TiO2. The width of the as-prepared TiO2 The high deposition temperature required for the CVD
nanobelt was about 72 nm. The growth direction was along the growth of TiO2 nanobelts should be attributed to the high
[100] direction of anatase crystal. melting point of Ti metal. Therefore, replacing metal Ti by
The hydrothermal method always involves three steps to pro- other materials, like TiCl4, can significantly decrease the dep-
duce TiO2 nanobelts — hydrothermal synthesis, acid treatment, osition temperature. This was probed by Shi and Wang, who
and calcination — which makes this process time consuming. produced TiO2 nanobelts at 650 °C using TiCl4 and H2O as the
However, the advantages of this method are its mass-produc- raw materials.[65]
tion potential, low cost, and easy control. Therefore, up to now, An alternative growth method is metallorganic chemical
most TiO2 nanobelts in literature have been synthesized by the vapor deposition (MOCDVD), which has several advantages —
hydrothermal method. in terms of compatibility with Si substrate and growth flexibility
of various materials — compared to wet-chemical approaches
and traditional CVD methods. MOCVD typically employs
3.2. Electrochemical Anodization organic titanium as the precursor, which allows reduction of the
growth temperature. For example, polycrystalline TiO2−δ nano-
The successful synthesis of TiO2 nanobelts through elec- belts with mixed rutile and anatase phase can be grown on bare
trochemical anodization without templates was reported by Si(100) substrate with [001] growth direction along at 510 °C
Chen and coworkers.[60] This method provides an efficient using (C11H19O2)2(C3H7O)2Ti as the metallorganic source.[52]
way to directly prepare TiO2-nanobelt-based photoelectrodes The as-grown TiO2−δ nanobelts were highly conductive.
for water splitting and biosensor applications. To conduct the The CVD method is generally used to synthesize rutile phase
electrochemical anodization, a standard two-electrode configu- TiO2 nanobelts and TiO2(B) nanobelts. Because of the difficulty
ration was employed, using platinum foil and titanium foil as in acquisition of the raw materials required, few works report
the cathode and anode, respectively. The electrolyte contained the synthesis of anatase TiO2 nanobelts by CVD. In addition,
ammonium fluoride, water, and ethylene glycol. By controlling though CVD has the advantage of yielding 1D nanostructures
the concentration of NH4F, the applied voltage, and the dis- with excellent electronic and optical properties of a large variety
tance between the two electrodes, anatase TiO2 nanobelts can of materials, it seems not to be the best method to produce
be synthesized on the surface of the titanium foil anode. The TiO2 nanobelts. This is because the morphologies of the TiO2
TiO2 nanobelts kept growing continuously when the anodiza- nanostructures currently obtained from CVD or MOCVD are
tion time was increased. irregular, obviously different from belt-like.
The electrochemical growth of the TiO2 nanobelts involves
three simultaneous reaction: electric field-assisted oxidation of
Ti metal to form TiO2, electric field-assisted etching of the oxide 3.4. Path-Directed Method
layer, and the chemical dissolution of TiO2 by fluoride.[61] It
should be pointed out that another key role of the fluoride ions Although the synthesis of TiO2 nanobelts is well developed,
is to control the growth of nanostructures with specific exposed the production of highly ordered TiO2 nanobelt arrays is still
crystalline facets.[44] Though this electrochemical method is challenging. 1D nanostructures with highly ordered patterns
simple, and the morphology of the formed TiO2 nanobelts can are desirable for patterned microelectronics. For example, Xia
be controlled by the experimental parameters such as applied and co-workers have fabricated TiO2 nanostructure arrays using
voltage and concentrations, these TiO2 nanobelts have very lithographically defined templates.[66] Kim et al. reported the
poor degrees of crystallinity. synthesis of well-ordered TiO2 nanostructures using surface
relief gratings on polymer films.[67] The polymer functioned as
the template in the sol–gel process.
3.3. Chemical Vapor Deposition Highly ordered TiO2 nanobelts with precisely controlled
width, thickness, and length can also be fabricated using a
The chemical vapor deposition (CVD) method, employing cata- simple, markless, path-directed method.[38] The path-directed
lytic substrates, has been developed as a convenient technique method of synthesis of TiO2 nanobelts involves three main
to obtain high-quality 1D nanostructures.[9,62] The CVD method steps. Firstly, a titanium film with a thickness of 40 nm is
is appropriate for the synthesis of 1D TiO2 on silicon substrates deposited on a clean glass substrate by electron-beam evapora-
coated with Ti, in the presence or absence of catalysts.[63] Typi- tion. Secondly, a writing process is applied to the Ti films using
cally, for the growth of 1D TiO2, a TiO2 seed layer must be a frequency-doubled Nd:YAG 532 nm laser. The high-energy

8 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
laser induces the oxidation of the irradiated Ti metal. Upon with the well-known performance of granulated TiO2 (P25).
Ti oxidation, the height of the oxidized regions increases due When the TiO2 nanobelts were manipulated into a nanopaper
to the interfusion of oxygen. Finally, a wet etching process is morphology, the nanopaper still had excellent photocatalytic
conducted in a dilute HF solution to remove the residual, performance, with about 60% degradation within 40 min
non-oxided Ti film, and the oxidized regions (TiO2 nanobelts) under UV light illumination. Note that the TiO2-nanobelt-
remain. This method results in polycrystalline TiO2 nanobelts. based nanopaper facilitates the recycling of the photocatalyst
during the photodegradation experiments, and that the slightly
lower photocatalytic performance can be easily compensated
4. Applications of TiO2 Nanobelts by increasing the number of layers within the nanopaper, as
the photocatalytic performance increases with the number of
TiO2 is a multifunctional material, widely used in photo- layers.
catalysis for water splitting and organic pollutant removal, solar The photocatalytic reactions proceeding on the surface of
cells, gas sensors, biosensors, and so on. TiO2 nanobelts, as one the TiO2 nanobelts involve both the hydroxyl radicals that come
of the most important 1D TiO2 nanostructures, have special from the reaction between valence band holes and chemisorbed
electronic and optical properties that bring about significant hydrogen groups or water molecules,[29] and also the superoxide
enhancement of their performance, and unique structural fea- radicals originating from reactions between the conduction
tures that widen their potential applications. In this section, we band electrons and the adsorbed oxygen molecules.[77] How-
summarize the basic applications of TiO2 nanobelts in photo- ever, for TiO2 nanobelts, the efficient photocatalytic degradation
catalysis, solar cells, gas sensors, biosensors, and electrochem- of methyl orange is mainly caused by superoxide radicals, and
ical energy storage devices. the hydroxyl radicals take a lesser role in the oxidation of the
methyl orange molecule. The better photocatalytic activity of
TiO2 nanobelts compared with TiO2 nanospheres is due to the
4.1. Photocatalysis underlying physical mechanisms. Firstly, TiO2 nanobelts have
greater charge mobility along the longitudinal dimension of
Since Fujishima and Honda reported the possibility of the crystals than that of TiO2 nanospheres. Secondly, there are
photocatalytic water splitting on a TiO2 electrode in 1972,[2] fewer localized states near the band edges and in the bandgap,
researchers have made numerous efforts to widen the due to the fewer unpassivated surface states in the TiO2 nano-
photocatalytic applications of TiO2 in the photodegradation of belts. Thirdly, the charge separation is enhanced due to trap-
various pollutants and the photogeneration of clean hydrogen ping of photogenerated electrons by chemisorbed molecular
fuel from water.[68] In one of the most significant works, it was oxygen on the (101) facet.[18] The following photoreactions typi-
demonstrated that TiO2 nanobelts have unique advantages cally represent the key steps of the formation of radicals,
over nanospheres in photocatalytic applications.[18] This work
clearly addressed the physical mechanism of the improved TiO2 + hv → TiO2 (h + + e − ) (4)
photocatalytic activity of TiO2 nanobelts over TiO2 nanospheres.
Very recently, our group has developed a hierarchical TiO2
nanowire/graphite fiber photoelectrocatalysis setup involving a TiO2 (e − ) + O2 → O2− (5)
wind-driven nanogenerator.[69]
In photocatalysis, the major and most important role of TiO2
is to transfer the energy from the light to the charge carrier. TiO2 (h + ) + H2O → OH⋅ (6)
Bulk TiO2 has a bandgap of about 3.2 eV and strong absorp-
tion in the UV, making it an ideal UV light photocatalyst. As a
result, TiO2 has the capability to efficiently photodegradate rho- Another important photocatalytic application of TiO2 nano-
damine B,[70] methylene orange,[71] and organic pollutants such belts is water splitting, for the photogeneration of hydrogen
as 2-mercaptobenzothiazole,[72] benzotriazole,[73] and 4-chloro- fuel.[32] For TiO2 nanobelts prepared from the electrochemical
phenol,[74] as well as to photocatalytically purify waste water anodization method, a maximum photoconversion efficiency
from the papermaking industry.[75] of up to 4.51% at an applied voltage of about 0.1 V (vs. SCE
However, as stated above, several problems still exist with (saturated calomel electrode)) can be reached, which is much
the use of granulated TiO2 photocatalysts, the main issue being higher than that of TiO2 nanotubes obtained using the same
recovery and recycling. Larger TiO2 nanobelts can well solve synthetic method (2.43% at 0.39 V).[78] Photoelectrochemical
the recycling problems, while having photocatalytic activity and (PEC) experiments have demonstrated that TiO2 nanobelts have
degradation efficiency comparable to granulated TiO2 photocat- more efficient separation of photoinduced charge carriers, com-
alysts.[32] Its 1D nanostructure allows easy collection from sus- pared with TiO2 nanotubes. The excellent PEC performance
pension for later reutilization, or its manipulation into a paper- should be associated with the long lifetime of the photoinduced
like morphology for continuous photodegradation of organic charge carriers, which has its origin in the efficient separa-
pollutants using a custom-made experimental set-up.[76] TiO2- tion, as well as a lower recombination rate.[60] For hydrogen
nanobelt-based nanopaper has good mechanical properties evolution using TiO2 nanobelts from alkali hydrothermal
and porosity. TiO2 nanobelts also show excellent photocatalytic process, hydrogen production rates up to 2.41 m−2 h−1 have
performance. They can almost completely degrade methylene been measured, higher than that of commercial TiO2 powder
blue within 55 min under UV light irradiation, comparable (P25, 1.91 m−2 h−1).[79]

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 9
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

4.2. Solar Cells stability and a large index of refraction, and is considered an
ideal candidate for optical biosensors.[6a,88] For example, glucose
TiO2 is also used for the fabrication of solar cells, particularly detection is still challenging, and numerous detection methods
dye-sensitized and organometallic halide perovskite solar cells. have been developed to realize fast and precise detection of glu-
In both cases, TiO2 plays the role of the electron transport layer. cose using biosensors.[89] Our group has systematically investi-
In dye-sensitized solar cells (DSSCs), a low cost alternative to gated the performance of TiO2-nanobelt-based electrochemical
mono- and polycrystalline silicon solar cells, TiO2 nanomate- biosensors. It was found that TiO2 nanobelts exhibited excel-
rials are photosensitized by dye molecules.[80] When TiO2 nano- lent electrocatalytic activity in the oxidation of nucleobases
particles are used, the weak contact between the neighboring (Figure 6b).[90] Two irreversible oxidation peaks appearing at
TiO2 nanoparticles, as well as defects on the nanoparticle sur- 0.62 V and 0.89 V can be clearly observed for the TiO2-nano-
faces significantly decrease the electron diffusion length.[28,81] belt-modified electrode. An even higher electrocatalytic perfor-
Highly crystalline 1D nanostructures usually have a lower mance can be obtained by hydrothermal treatment of the TiO2
trap density and offer more direct paths for charge transpor- nanobelt in diluted HCl aqueous solution. It is suggested that
tation, displaying lower electrical resistances for charge trans- this is related to the increased specific surface area of the acid
port and much longer electron diffusion lengths of over tens of corroded TiO2 nanobelt.
micrometers.[82] These properties make them ideal candidates Taking advantage of the excellent electrocatalytic oxida-
for DSSCs. For example, solar cells using single crystalline tion of nucleobases using TiO2 nanobelts, a biosensing elec-
anatase TiO2 nanorods from a hydrothermal method display trode employing acid corroded TiO2 nanobelts was rationally
a high photon-to-current conversion efficiency of 7.29%.[83] designed to selectively detect the match and mismatch of single
Feng et al. used vertically aligned single crystal rutile TiO2 nucleobases.[91] As a result, the TiO2 nanobelts displayed dis-
nanowires to assemble DSSCs with an open circuit voltage of tinguished selectivity and sensitivity. Figure 6c shows the elec-
0.78 V, a fill factor of 0.68 and a short-circuit photocurrent den- trochemical responses of single base-pair interactions at the
sity of 6.95 mA cm−2, obtaining a solar conversion efficiency of TiO2 nanobelts modified electrodes in PBS (phosphate buffer
3.68%.[84] solution) at pH 7.4. Obviously, the surface-coarsened TiO2
Like nanorods and nanowires, TiO2 nanobelts are also ideal nanobelts (CTNs) can selectively detect the perfect match or
candidates for DSSCs, and even exhibit better performance mismatch of single nucleobases. It is shown that the oxidation
than the other 1D nanostructures. Kai Pan et al. have compared peak potential has a negative shift of 30 mV for G (guanine)–U
the performance of DSSCs using TiO2 nanobelts and TiO2 (uracil) mismatch and G–T (thymine) mismatch, compared to
nanorods.[85] Under similar conditions and a using a xenon the perfect match of G–C (cytosine). Similar phenomena can be
lamp with intensity 50 mW cm−2, the DSSC made from TiO2 observed for the A (adenine)–C mismatch, A–U perfect match
nanobelts shows a better photoelectrical energy conversion effi- and A–T perfect match. Based on the excellent sensitivity and
ciency of 2.2% than those made from TiO2 nanorods (1.8%). selectivity of TiO2-nanobelt-based electrochemical biosensor,
In addition, the TiO2 nanobelt exhibited a bigger photovoltage TiO2 nanobelts were also used to investigate cancer cells and
than that of TiO2 nanorods cells, indicating that the conduc- anticancer drug effect (Figure 6d).[92]
tion band level of TiO2 nanobelts was more negative and more
adjacent to the lowest unoccupied molecular orbital (LUMO)
of the dye molecule. As another example, the IPCE (incident 4.4. Gas Sensors
photon-to-electron conversion efficiency) of DSSCs based on
TiO2 nanobelts produced from hydrothermal synthesis exhibits Metal oxide semiconductor gas sensors are based on the varia-
a maximum at 530 nm, with a conversion efficiency of 33.2%, tion of electrical resistance of a porous film upon gas molecules
under illumination of AM 1.5 G (100 mW cm−2).[27] There is no adsorption at the surface.[93] The electrical resistance of the sem-
doubt that TiO2 nanobelts will attract more attention for DSSC iconductor sensor depends upon the type and concentration
applications as the understanding of their electrical proper- of the adsorbed gas, which is the basic mechanism providing
ties is improved. Although, to the best of our knowledge, TiO2 selectivity.[94] For practical applications, the sensing material
nanobelts have not yet been used to replace TiO2 nanoparticles should fulfill particular requirements critical to the operation
in the fabrication of perovskite-based solar cells, they may also of real sensors, including sensitivity, selectivity, and speed of
be able to provide advantages in this emerging type of photo- response requirements.[95] Among the numerous metal oxides,
voltaic device. TiO2 is a particularly interesting gas sensor material, which has
proven excellent performance in detecting different species,
such as O2,[96] H2S,[97] H2,[98] CO,[99] NO2,[100] NH3,[5,101] and so
4.3. Biosensors on. Besides, 1D nanostructures with high surface-to-volume
ratio, are especially suitable in various gas sensor architectures,
As stated above, TiO2 nanobelts are highly stable and have good such as nanowire-based films, nanowire arrays, and even single
compatibility with biomaterials. At the same time, they are able nanowire sensors.[102]
to enhance enzyme catalytic performance.[86] These merits posi- Our group has fabricated gas sensors using nanostructured
tion TiO2 as an excellent candidate for various biosensors. In sheets based on anatase TiO2 nanobelts (Figure 7a).[17] It was
addition, good electronic properties and high surface to volume found that TiO2 nanobelts showed good response to O2. With
ratio ensure the fast transportation of charge carriers, which increasing O2 concentration, the resistance increased almost
guarantees a high sensitivity.[87] TiO2 has excellent aqueous linearly from about 3.9 MΩ at 100 ppb to 5.4 MΩ at 1000 ppb.

10 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW

Figure 6. a) SEM image of TiO2 nanobelts. Reproduced with permission.[90] Copyright 2011, Royal Society of Chemistry. b) Cyclic voltammograms of
various chemically-modified electrodes in 0.1 M PBS (pH 7.4) without and with 0.1 mM guanine (G) and adenine (A). Potential sweep rate 100 mV s−1.
TNs/CA/GCE refers to the TiO2-nanobelt-modified glassy carbon electrode. CTNs/CA/GCE refers to the glassy carbon electrode modified by surface-
coarsened TiO2 nanobelts. Reproduced with permission.[90] Copyright 2011, Royal Society of Chemistry. c) SWV of single base-pair matches: TNs/
CA/GCE and CTNs/CA/GCE in 0.1 M PBS (pH 7.4) containing 0.1 mM nucleobases. Reproduced with permission.[91] Copyright 2011, Royal Society
of Chemistry. d) Cyclic voltammetry (CV) curves for various modified electrodes in 0.1 M PBS (pH 7.4) with and without 0.1 mM O6-benzylguanine
(O6BG), and the magnified cyclic voltammogram of the partial area in, sweep rate, 100 mV s−1. Reproduced with permission.[92] Copyright 2013, Royal
Society of Chemistry.

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 11
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

Figure 7. a) Low-magnification SEM image of the TiO2 nanobelt paper and high-magnification image of a small area of the same paper. Reproduced
with permission.[17] b) Response of nanostructured TiO2 sheet to different oxygen concentrations: resistance–concentration curve and gas sensitivity
curve of TiO2 sheet at room temperature, as well as the sensor response to O2 at different concentrations. Reproduced with permission.[17] c) (left)
Response curves and (right) sensitivity profiles of ethanol vapor sensors based on TiO2 nanobelts, Ag–TiO2 nanobelts, surface-coarsened TiO2 nano-
belts, and surface-coarsened Ag–TiO2 nanobelts upon exposure to different concentrations of ethanol vapor at 200 °C. Reproduced with permission.[103]
Copyright 2010, American Chemical Society.

At room temperature, detection levels down to 100 ppb, with sensing mechanism is based on the variation of concentration
still-high resistance variations of up to one order of magni- of surface oxygen vacancies in the TiO2 nanobelts with the
tude, was accomplished. The good O2 sensing performance oxygen partial pressure. This change in the concentration of
of TiO2 nanobelts is due to the high density of surface sites oxygen vacancies translates into a modification of the concen-
available for gas adsorption and the high influence these tration of trapped electrons, and thus into a variation of the
sites have on the overall electrical properties of the material charge carrier density, measured as a change in the electrical
because of the high surface-to-volume ratio. The suggested resistivity.

12 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

TiO2@Co3O4 and TiO2@MoS2 has been proposed.[109e,112] The

REVIEW
As expected, TiO2 nanobelts also exhibit excellent perfor-
mance in the sensing of other target gases like ethanol vapor, goal of the shell, made of a high capacity material, is to facili-
which leads to potential application in breath alcohol tests. TiO2 tate and increase Li uptake, while the TiO2 nanobelt provides
nanobelts show a rapid response to ethanol gas, even at con- a suitable and mechanically stable framework that provides
centrations as low as 20 ppm.[103] This sensing performance both a high electrode–electrolyte interface and efficient charge
can be further improved by hydrothermally treating the TiO2 transport.
nanobelts in a diluted acid solution, which creates surface To improve performance in LIBs, particularly with regard to
coarsened TiO2 nanobelts, or by assembling noble metal nano- electrical conductivity, TiO2 nanobelts can be also wrapped with
particles on the surface of the TiO2 nanobelts. Metal nanoparti- graphene or other forms of carbon, which act as a highly effi-
cles can lower the gas sensing temperature of TiO2 nanobelts, cient porous current collector.[109a,b,e] Moreover, very recently,
and the surface coarsening can improve the sensitivity of TiO2 a novel hybrid Li-ion capacitor combining a graphene-based
nanobelts. Ag nanoparticle assembled on surface-coarsened electrochemical double layer capacitor-type cathode with a TiO2
TiO2 nanobelts display a significantly improved sensing perfor- nanobelt based LIB-type anode has been demonstrated to have
mance, which is much better than many other sensors based high energy and power densities.[113]
on ZnO nanowires,[104] ZnO nanopillars,[105] SnO2 nanorods[106]
and Pt–TiO2 nanowires arrays.[107]
5. Modification of TiO2 Nanobelts
4.5. Energy Storage Devices Pristine TiO2, the first-generation material, has been employed in
photocatalytic degradation and mineralization of a large number
TiO2 nanostructures, and particularly TiO2 nanobelts, have of organics, as well as the production of clean hydrogen fuel
also been used in the field of energy storage as supercapacitor from water. Unfortunately, as mentioned above, TiO2, either in
electrodes,[108] and especially as anode materials in lithium rutile or anatase phase, has a wide bandgap resulting in a photo-
ion batteries (LIBs).[109] The advantages of TiO2 are its low threshold that extends from the UV region to about 400 nm. This
cost, mechanical stability, environmental friendliness, safety, inherent limitation of TiO2 curtails the efficiency of conversion of
relatively high lithium storage capability, and good cyclability. the solar energy that reaches the earth’s surface. Like bulk TiO2,
As a supercapacitor electrode material, TiO2 nanobelts shows TiO2 nanobelts also have a wide bandgap and can only absorb
improved energy storage performance, including significant UV light, which imposes a major challenge for solar energy
enhancement of specific capacitance, rate capability, energy conversion. Numerous efforts have been made to overcome this
density, and power density, when coupled with reduced gra- challenge. These efforts can typically be divided into two dif-
phene oxide (rGO), due to their unique shape, better charge ferent approaches. The first approach is the modification of TiO2
transport properties, and larger area of contact with the rGO nanobelts with heteroatoms, including cations and anions. The
nanosheet compared with TiO2 nanoparticles.[108a] This work second is the modification of TiO2 nanobelts by constructing a
clearly demonstrates that it is necessary to consider the shape secondary phase on the TiO2 nanobelt surface, to form a sur-
and coupling effects of metal oxide and carbon materials for face heterostructure. In this section, we review the modification
supercapacitors. All main TiO2 polymorphs have been proven of TiO2 nanobelts based on the above two routes, summarizing
to be excellent anode materials for LIBs. Among them, the the achievements, and providing an overview on solutions to the
anatase phase is found to have the best properties, with a theo- challenges associated with the use of TiO2 in photocatalysis.
retical capacity of 335 mA h g–1, a flat operating potential and
a low volume expansion, <4%, during Li+ charge/discharge
cycles.[109f,110] However, excellent capacities and cycleabilities 5.1. Modification with Heteroatoms
have been also obtained for TiO2(B) nanobelts,[109a,111] and also
for Magnéli-phase TinO2n−1 nanobelts.[109c] The main drawbacks It is thought that introducing heteroatoms into the crystalline
of TiO2 are its relatively poor Li+ and electronic conductivities. structure of the TiO2 nanobelt can reduce the photothreshold
The use of nanostructured films allows for improvements in energy by extending the light absorption edge to visible light.
the electrode/electrolyte interface, and thus leads to reduction Numerous attempts toward achieving this goal have been made,
in the Li+ transport path within the electrode. This has been employing various approaches including hydrothermal,[114]
experimentally demonstrated for rutile TiO2 by measuring sol–gel,[115] electrochemical methods,[116] and so on. The het-
increasingly higher capacities, up to 378 mA h g−1, when ero-elements introduced into TiO2 include not only anions
reducing the TiO2 crystal domain size down to 15 nm. (non-metal doping) but also the cations (metal doping). The
However, the use of nanoparticles further reduces the elec- following section aims to address the corresponding changes
tronic conductivity of the electrode. On the other hand, the use of in electronic properties and optical properties originating from
nanobelts allows improvements in the electron transport proper- the introduction of anions and cations.
ties of the material, while still providing a network of electrolyte
channels for efficient Li+ transport. Moreover, the nanobelt geom-
etry maintains the high interface area with the electrolyte and the 5.1.1. Modification with Cations
excellent ability to accommodate lithiation-induced stresses.[110]
To further improve the performance of TiO2-nanobelt-based Cation doping is usually realized by introducing transition
LIBs, the use of coaxial nanobelts heterostructures such as metal ions, such as V,[117] Fe,[118] Co,[114a,119] and Cr[120] and so

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 13
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

Figure 8. a) SEM and HRTEM image of Bi doped TiO2 nanobelt, as well as the corresponding UV–vis spectra. Reproduced with permission.[114b]
Copyright 2009, Elsevier. b) SEM image, HRTEM image of the RE doped TiO2 nanobelt, and upconversion (UC) fluorescence emission spectra of the
RE-doped TiO2 nanobelts (λex = 980 nm): (left) the Er–TiO2-01 and Er–TiO2-02; (right) the RE–TiO2-01 and RE–TiO2-02 (inset is the magnified spec-
trum around wavelength 660 nm). For comparison, the UC emission spectra of the Er–LTO-01 and RE–LTO-01 as the precursors of the Er–TiO2-01 and
RE–TiO2-01 are also listed in the figure. LTO referes to layered titanate nanobelts. The number (01 and 02) refers to the concentration level of dopant.
Reproduced with permission.[121] Copyright 2010, Elsevier.

on, into the crystalline phase of TiO2. The most commonly metal atoms into the TiO2 nanobelt does not change the crystal
utilized method for the modification of TiO2 nanobelts with structure and the morphology. However, even a tiny amount of
cation heteroatoms is the hydrothermal method.[114c,d,121] heteroatoms can trigger significant changes. For example, com-
The cation modification of TiO2 has shown both positive and pared to the undoped TiO2 nanobelts, the Bi-doped TiO2 nano-
negative effects on the photocatalytic activity of TiO2.[122] The belts show a red-shift of the absorption edge, and have strong
positive aspect is that cation modification can decrease the responses to visible light irradiation, as shown in Figure 8a.[114b]
photothreshold energy of TiO2, while the negative aspect is Another interesting group of dopants are the rare earth (RE)
that the introduced metal ions act as new recombination center metals. Rare earth ions are known to have excellent upcon-
for electrons and holes.[48] For example, Fe, Mo, Ru, Os, Re, V, version ability.[126] Upconversion refers to optical processes
and Rh atoms significantly increased the photoreactivity for featuring the successive absorption of two or more photons
both oxidation and reduction, while Co and Al decreased the through intermediate, long-lived energy states, followed by
photoreactivity, which is largely dependent on the concentra- the emission of output radiation at a shorter wavelength than
tion, energy state, electron configuration and distribution of the the incident wavelength.[127] This effect can be introduced into
heteroatoms.[123] TiO2 nanobelts by doping with rare earth ion such as Er3+ and
TiO2 nanobelts have a layer structure which is beneficial for Yb3+.[121] According to results of experiments investigating
the introduction of heteroatoms. For example, lithium titanate upconversion fluorescence emission, compared with undoped
nanobelts are an ideal anode material for high performance TiO2 nanobelts, Er-doped TiO2 nanobelts showed two dif-
Li-ion batteries, as discussed above.[109c] To date, modification of ferent emission peaks under excitation illumination of wave-
TiO2 nanobelts with metal ions is limited to the hydrothermal- length 980 nm at room temperature. The two emission peaks
process-assisted ion-exchange method. Most current literature are a strong green emission and a relatively weak red emis-
is focused on the introduction of Bi,[114b] Co,[119,124] Ni,[114a] sion. Other results show that by modifying the TiO2 nanobelts
Fe,[118,123] and rare earth metal ions.[121,125] With the introduc- with rare earth ions, 980 nm light can be converted into two
tion of different metal ions into the crystalline structure of different wavelength ranges of 520–570 nm and 630–680 nm
TiO2 nanobelts, different properties are usually expected. For (Figure 8b).[121] This upconversion effect may be used to realize
example, Co or Ni doped TiO2 nanobelts displayed room tem- the preparation of near-infrared photocatalysts which can
perature ferromagnetism, while Bi doped TiO2 nanobelts can largely enhance the utilization and conversion efficiency of
extend the absorption edge to the visible light range to enhance solar light energy. However, more research efforts are needed,
the light harvesting ability.[114a,124] Generally, introducing hetero because introducing different metal atoms usually brings about

14 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
Figure 9. a) Various schemes illustrating the possible changes that might occur to the bandgap electronic structure of anatase TiO2 on doping with
various non-metals: 1) bandgap of pristine TiO2; 2) doped TiO2 with localized dopant levels near the VB and the CB; 3) bandgap narrowing resulting
from broadening of the VB; 4) localized dopant levels and electronic transitions to the CB; and 5) electronic transitions from localized levels near the
VB to their corresponding excited states for Ti3+ and F+ centers. Ti3+ and F+ refer to the color centers which can absorb visible light radiation. Repro-
duced with permission.[134] Copyright 2006, American Chemical Society. b) Diffusive reflectance spectra of pristine and nitrogen doped anatase TiO2
nanobelts, and electronic density of states (EDOS) for TiO2 nanobelts. Reproduced with permission.[135] Copyright 2009, American Physical Society.
c) UV–vis absorption spectra of N–F co-doped TiO2 nanobelts. Reproduced with permission.[128] Copyright 2012, American Chemical Society. d) UV–vis
absorption spectra of as-prepared N–S co-doped TiO2 nanobelts. Reproduced with permission.[129] Copyright 2011, Elsevier.

different properties to the TiO2 nanobelts. Due to this com- belts.[128,129,122a,134] Determining the in-depth mechanism of vis-
plexity, there are no general rules for metal ion modification of ible light photocatalytic activity of anion doped TiO2 has proven
TiO2 nanobelts. controversial. Wu’s group has conducted systematic investiga-
tions on the N-doped TiO2 nanobelts. It is found that N-doping
of TiO2 nanobelt results in the presence of mid-gap states in
5.1.2. Modification with Anions the energy band of the TiO2 nanobelt, rather than narrowing
of the bandgap. N-doping of TiO2 nanobelts can result in both
An alternative approach for lowering the photothreshold of TiO2 positive and negative effects on the photocatalytic activity of
is to modify the TiO2 nanobelts by doping with anions. The TiO2 nanobelts, i.e., limited visible light photocatalytic activity
modification strategies include the solvothermal method,[128] as well as the introduction of mid-gap energy states as charge
hydrothermal method[129] and high temperature annealing in recombination centers.
specific gas atmosphere.[130] Since the first report of the vis- Usually, TiO2 nanobelts modified with anion heteroatoms
ible light activity of N doped TiO2,[131] it has been found that are characterized by a change in color. For example, N-modified
N-doped TiO2 shows extended light absorption behavior from anatase TiO2 nanobelts display a pale yellow color compared to
the UV to the visible light regions, leading to the visible light the white color of pristine TiO2 nanobelts. This color change typ-
photocatalytic activity of TiO2. However, the fundamental ically originates from the red-shift of the light absorption edge
aspects of the origin of the visible light photocatalytic activity of the TiO2 nanobelt. This red-shift effect of the anion modi-
of the N-doped TiO2 were still unclear until the work of Wu’s fied TiO2 nanobelts is very common, as shown in Figure 9b.[135]
group.[132] Actually, since the early stages of research into anion- The advantage of this effect is the extension of light absorption
doped TiO2, it has been well accepted that introducing anions range from the UV into to the visible light range, increasing
usually results in a bandgap narrowing or introduces interband the utilization efficiency of solar light.[130] Meanwhile, the
levels, as shown in Figure 9a.[133] These positive effects of anion introduced anion atoms usually have two possible sites in the
modification of TiO2 have already been probed for TiO2 nano- crystalline structure of TiO2 nanobelt, i.e., interstitial sites and

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 15
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

substitutional sites. As a result, oxygen vacancies are usually and pits on the TiO2 nanobelt surface through an in situ dis-
created with anion modification. Varying the concentration of solution–precipitation process. In addition, the acid corrosion
the anion heteroatoms can change the related concentrations of method results in intimate contact between the in situ created
the interstitial atoms and oxygen vacancies. nanoparticles and the TiO2 nanobelts, facilitating charge-carrier
Aside from N, other elements like C, S, F, can also be intro- separation and transfer.
duced into TiO2 nanobelts. However, though there exist few Broaden the Light Absorption Range: As previously discussed
reports on the doping of TiO2 nanobelts with just C, S, or F, extensively, TiO2 nanobelts have a wide bandgap and can only
the existing literature is mainly focused on the co-doping absorb UV light, leaving the majority of solar light energy
of TiO2 nanobelts with N and F, or N and S, as shown in wasted. To solve this issue, two different routes can be consid-
Figure 9c and 9d.[129] These co-doped TiO2 nanobelts display ered. One route is optimization of the surface and bulk phys-
obvious extension of their optical absorption edge from UV to ical and chemical properties by introducing oxygen vacancies,
visible light regions. However, theoretical analysis of the co- tuning crystal phases, and/or modification with hetero-atoms.
doping of TiO2 nanobelts has thus far never been reported. Another route is the construction of surface heterostructures
by introducing narrow bandgap semiconductors, dye mol-
ecules, and up-converting materials at the TiO2 nanobelt sur-
5.2. Modification with Surface Heterostructures face. The former route is typically based on bandgap narrowing
and the introduction of interband energy levels in the bandgap
Though TiO2 nanobelts have several improved properties com- of TiO2. However, this route usually leads to some negative
pared to TiO2 nanoparticles, such as better electronic properties, effects on the resulting photocatalyst, such as poor stability
mechanical properties and the ability to be easily recycled and and even worse charge separation. The latter route is based on
reutilized as photocatalysts, they still suffer from the aforemen- a synergistic effect between the TiO2 and the secondary phase
tioned wide bandgap, which limits their applications to the UV in the heterostructure. In fact, this route has shown tremen-
light range, and from relatively fast charge-carrier recombina- dous advantages and can be split into two different strategies,
tion, which reduces the effective number of electrons and holes depending on the range of light absorption of the secondary
for degradation reactions. phase.
Ideally, for photocatalytic applications, the structure should The modification of TiO2 nanobelts with visible-light photo-
make full use of the solar light spectrum, from UV to visible catalysts to form surface heterostructures can extend the light
light, and even infrared light. The material should also have the absorption into the visible. Semiconductors with a bandgap
ability to effectively separate the photoinduced charge carriers, energy lower than 3.2 eV, i.e., a cut-off edge of light absorp-
to increase solar light conversion efficiency. However, presently, tion above 400 nm, can be used as potential visible light
no individual photocatalyst has been found that can meet the absorbers.[137] Transition metal oxides, transition metal chalco-
above criteria. One promising protocol to solve these problems genides, and their corresponding composite oxides and chalco-
is the construction of surface heterostructures on TiO2 nano- genides are potential materials for consideration.[138]
belts. In this section, we systematically summarize the design The modification of TiO2 nanobelts with near-infrared light
principles for the construction of TiO2 nanobelt surface hetero- absorbers can result in materials covering nearly the whole of
structures. Based on these design principles, several kinds of the solar light spectrum. These kind of surface heterostructures
TiO2 nanobelt surface heterostructures that displayed enhanced should significantly increase the light conversion efficiency,
photocatalytic activity have been successfully synthesized. because infrared light accounts for a significant part of the
solar light energy.
One alternative and effective method to use near-infrared
5.2.1. Design Principles of TiO2 Nanobelt Surface Heterostructures light is to load upconversion materials onto TiO2 nanobelts.
The upconversion of infrared light to visible light usually hap-
The construction of rationally designed, high performance TiO2 pens in materials containing d-block or f-block elements.[139]
nanobelt heterostructures should obey several basic design The most famous upconversion material is NaYF4 doped with
principles in order to control fundamental aspects such as rare earth ions.[140] However, it is not possible to make full use
charge separation, interfacial charge transfer, and electronic of the infrared light by means of upconversion, because the
properties of materials. excitation of the upconverting materials is limited to a single or
Enlarge the Active Photocatalytic Surface Area: Photocatalytic few wavelength of light. Presently, there is a large requirement
reactions occur at the active sites on the surface of a photo- for the development of more efficient near-infrared or infrared
catalyst. Thus, the number of active sites directly influences the light photocatalysts.
performance of the photocatalyst. The most common method Surface Plasmon Resonance Effects: TiO2 nanobelts are UV-
for increasing the number of active sites is to increase the TiO2 light photocatalysts with strong absorption ability of UV light.
nanobelt surface area by depositing nanoparticles onto its sur- However, there is still room for further enhancement of the
face. However, some drawbacks still exist, such difficulties in UV-light photocatalytic activity. One effective way is to use the
controlling particle size, distribution uniformity, and density plasmon resonance effect originating in noble metals, such as
and a poor connectivity between the nanoparticle and the TiO2 Au,[141] Ag,[142] and their alloys with Pt and Pd.[143] The solar
nanobelt substrate. One very effective way to increase the sur- energy conversion efficiency of semiconductor–metal hetero-
face area of TiO2 nanobelts is the application of hydrothermal junctions can be improved by the transfer of plasmonic energy
acid corrosion,[136] which can create numerous nanoparticles from the plasmonic metal to the semiconductor via three

16 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

underlying mechani[Link] i) photonic enhancement,[145]

REVIEW
nanobelt surface heterostructures, and conducting polymer–
ii) direct hot electron transfer,[146] and iii) plasmon-induced res- TiO2 nanobelt surface heterostructures.
onance energy transfer (PIRET).[147] So far, direct hot electron Metal–TiO2 Nanobelt Surface Heterostructures: Noble metal
transfer processes have shown limited contribution to solar nanoparticles can be assembled onto the surface of TiO2 nano-
energy conversion efficiency in semiconductor–metal hetero- belts to form surface heterostructures, including Schottky junc-
junctions. In contrast, PIRET has been reported to enhance the tions and plasmon-resonance-effect structures. Because of the
solar energy conversion of semiconductors remarkably.[148] excellent control over the electron transfer processes in storing
Schottky Junctions: When a metal and a semiconductor and shuttling photogenerated electrons, noble metals such as
come in to contact, an energy barrier is formed at the interface Au,[152] Ag,[153] Pt, and Pd[154] have attracted much attention and
between them if the Fermi level of the metal is below that of have been coupled with various TiO2 nanostructures.[155] In the
the semiconductor. Such a structure is is termed a Schottky heterostructure, the TiO2 nanobelt acts not only as a photocata-
junction.[149] The main feature of the Schottky barrier is the lytic active platform, but also provides a semiconductor band to
Schottky barrier height which, to a first approximation, is the couple with the metal to form a Schottky barrier.[156] The noble
energy difference between the work function of the metal metal nanoparticles can either be grown in situ or immobi-
and the electron affinity of the semiconductor. By reasonable lized on the surface of the TiO2 nanobelt.[132,133] In situ growth
manipulation, the barrier can be engineered such that that typically proceeds through photoreduction[103,155] or chemical
photoinduced charge carriers in the semiconductor do not reduction in the presence of reducing agents,[157] while immo-
accumulate at the interface but flow into the metal, leaving bilization usually involves the utilization of specific functional
the holes in the valence band of the semiconductor, and hence groups as linking agents. The key point for fabrication of such
achieving the goal of charge carrier separation.[150] a heterostructure is control of the size and distribution of the
p–n Junctions: When a p–n junction is formed at the interface noble metal nanoparticles.
between two types of semiconductor, a p-type and an n-type, a The synthetic process involves the photoreduction of the
local electric field appears across the space charge region.[151] noble metal salt in aqueous solution. The deposited noble metal
Driven by the electric field, the photogenerated electrons move nanoparticles can improve both the physical and chemical per-
towards the n-type semiconductor, and holes flow into the formance of the photocatalyst. Consequently, the surface energy
p-type semiconductor, accomplishing an effective separation of structure of the TiO2 nanobelt and the recombination dynamics
photoinduced charge carriers. TiO2 nanobelts are n-type semi- of photogenerated electrons and holes can be adjusted to
conductors, and hence p-type semiconductors, such as NiO, enhance the photoelectronic properties. These positive effects
Ag2O, Cu2O, can be coupled onto the surface of TiO2 nanobelt have been demonstrated in the excellent gas sensing perfor-
to form the desired p–n junction. mances and photocatalytic activities. With delicate control over
Band Structure Matching: Photogenerated electrons and the experimental parameters, the noble metal nanoparticles
holes can be effectively separated through the use of p–n and can be uniformly distributed on the surface of TiO2 nanobelt
Schottky junctions. However, these two types of heterostruc- — rather than forming a core–shell nanostructure — as shown
tures have some disadvantages. Firstly, in most of the present in Figure 10.[103,155,158] By taking advantage of photoreduction,
studies the metals used for photocatalysis are precious and both the Ag and Pd nanoparticles can be grown with intimate
rare, which increases the potential cost. Secondly, most semi- contact with the TiO2 nanobelt, thus having excellent charge-
conductors used in photocatalysis are n-type, and few of the carrier transfer properties and improved photocatalytic activity.
possible p-type semiconductors are suitable for photocatalysis. At the same time, the Ag nanoparticle–TiO2 nanobelt surface
In addition, the direct preparation of p-type semiconductors is heterostructure and the Pd nanoparticle–TiO2 nanobelt surface
not easy. Fortunately, it is still possible to construct TiO2-nano- heterostructure display excellent gas sensing performance with
belt-based heterostructures as long as the band structure of the very high sensitivity to ethanol gas. Pt nanoparticles can also
semiconductor matches that of TiO2. The band structure of the be assembled onto the surface of TiO2 nanobelts to form Pt
secondary semiconductor should be such that it forms a type-II nanoparticle–TiO2 nanobelt surface heterostructures, as shown
alignment that facilitates the transfer of electrons and/or holes by the formation of Pt nanoparticles with a mean size of about
between the two semiconductors. In this way, the range of 3.3 nm, homogeneously distributed on the surface of TiO2
available semiconductors suitable for the construction of useful nanobelt.[159] These noble metals are found to form a bimetallic
heterostructures with TiO2 nanobelts can be widened. phase on the TiO2 nanobelt to form the surface heterostructure.
For example, Au and Pd can form bimetallic alloy nanoparticles
on TiO2 nanobelts for enhanced photocatalytic aerobic oxida-
5.2.2. Classification of TiO2 Nanobelt Surface Heterostructures tion of benzyl alcohol.[160]
Metal Oxide–TiO2 Nanobelt Surface Heterostructures: Metal
According to the design principles stated in Section 5.2.1, and oxide–TiO2 nanobelt surface heterostructures can be divided
the different compositions of the second phase on TiO2 nano- into three different types according to the light absorption
belts, TiO2 nanobelt surface heterostructures can be classi- properties of the metal oxide nanoparticle: heterostructures
fied into several categories. In this section, the composition, formed by UV light-active metal oxide nanoparticles and TiO2
microstructures, and properties of different categories of TiO2 nanobelts; heterostructures formed by visible light-active metal
nanobelt surface heterostructures are summarized, including oxide nanoparticles and TiO2 nanobelts; and heterostructures
metal–TiO2 nanobelt surface heterostructures, metal oxide– formed by near-infrared or infrared light-active metal oxide
TiO2 nanobelt surface heterostructures, chalcogenide–TiO2 nanoparticles and TiO2 nanobelts. From the point of view of

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 17
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

Figure 10. a) SEM and TEM images of Ag nanoparticle–TiO2 nanobelt surface heterostructure. SEM image: Reproduced with permission.[103] Copy-
right 2009, American Scientific Publishers; TEM images: Reproduced with permission.[158] 2010, American Chemical Society. b) Pd nanoparticle–TiO2
nanobelt surface heterostructure. Reproduced with permission.[155] Copyright 2012, Elsevier. c) TEM images of Pt nanoparticle–TiO2 nanobelt surface
heterostructures. Reproduced with permission.[159]

band structure and carrier separation, these metal oxide–TiO2 obvious enhancement of the photocatalytic performance of the
nanobelt surface heterostructures are based on several of the heterostructure is ascribed to the enlarged specific surface area,
design principles mentioned previously. These surface hetero- increase in the number of active sites, and the heterostructure
structures can be tuned to optimize charge transfer processes formed between the TiO2 nanoparticle and TiO2 nanobelt. It is
and extend the light absorption to the visible light and the near- notable that the TiO2 nanoparticle and the TiO2 nanobelt have
infrared light regions. intimate contact, which is beneficial for the interfacial charge
To fabricate metal oxide nanoparticle–TiO2 nanobelt transfer.
surface heterostructures, many methods have been devel- By using acid–corroded TiO2 nanobelts as a substrate, our
oped, including sol–gel,[150,151] solution decomposition pro- group has developed several TiO2 nanobelt-surface heterostruc-
cesses,[152,153] wet chemical deposition,[161] hydrothermal pro- tures. The secondary materials of these surface heterostruc-
cesses,[162] electrodeposition,[163] and so on. To achieve efficient tures include Ag2O nanoparticles,[33] NiO nanoparticles,[164]
function of the surface heterostructure, two key points must be CeO2 nanoparticles,[165] and Cu2O nanoparticles,[166] as well as
taken into consideration. The first is control of the size of metal metal oxide compound Ag3PO4 nanoparticles.[167] These sur-
oxide nanoparticles and their distribution on the surface of face heterostructures display enhanced UV light photocatalytic
the TiO2 nanobelts. The second is the use of eco-friendly tech- activity, even better than that of P25, and excellent visible light
niques. That is to say, the synthetic method should be easily photocatalytic performance. The common aspects of these sur-
scalable, and involve no toxic materials. face heterostructures lie in the following: i) the secondary mate-
As an important basic work regarding the modification of rial has intimate contact with the TiO2 nanobelt. For example,
TiO2 nanobelts to improve the photocatalytic activity, our group Ag2O nanoparticles can be assembled onto the surface through
developed an effective approach to modify the TiO2 nanobelts electrostatic reaction between the negatively charged TiO2
in situ, to fabricate TiO2 nanoparticle/TiO2 nanobelt hetero- nanobelts and positively charged Ag2O nanoparticles. ii) The
structures based on an acid–hydrothermal process.[136] The secondary phase in the heterostructure has uniform distribu-
morphology and photocatalytic performance in the degradation tion and appropriate coverage which does not prevent the
of organic molecules using TiO2–nanoparticle/TiO2–nanobelt absorption of UV light by the TiO2 nanobelt. iii) Enhanced
surface heterostructures depend on the acid corrosion condi- charge separation due to the interface formed between the two
tions. With an increase of acid corrosion time, the photocata- parts of the heterostructure. For example, NiO is a p-type semi-
lytic activity of the heterostructures gradually increases. The conductor and TiO2 is an n-type semiconductor. Through the

18 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
in situ decomposition by high temperature calcination, a p–n Quantum dots have the advantage of band structure manipu-
junction can be formed between the NiO nanoparticles and lation, while layered 2D chalcogenides like MoS2 are famous
TiO2 nanobelts.[164] The NiO–TiO2 nanobelt p–n junction also for their high electrolytic activity due to the numerous active
shows excellent biosensing performance.[168] iv) The secondary sites on the edges.[178] Recently, a 2D MoS2 layer was synthe-
part usually has high activity under visible light irradiation. It sized in situ on the surface of a TiO2 nanobelt to form a novel
should be pointed out that the surface heterostructure formed 2D–1D surface heterostructure (Figure 11b).[176] The 2D MoS2
of CeO2 nanoparticles@TiO2 nanobelts shows a different nanosheet–1D TiO2 nanobelt surface heterostructure has
photocatalytic mechanism, the capture–photodegradation– increased contact area and intimate contact because of the
release mechanism.[165] In this mechanism, the organic special in situ grown nanostructure. Fast electron transport
molecules were captured by CeO2 nanoparticles, degraded by and interfacial transfer were expected in this novel surface
the photoinduced free radicals, and then released to the solu- heterostructure, ensuring high photocatalytic performance for
tion. The CeO2/TiO2 nanobelt surface heterostructure has high water splitting for hydrogen evolution and photodegradation
degradation efficiency and broad active light wavelength, as of organics. Similarly, the amount of MoS2 loaded on the TiO2
well as good stability. nanobelt surface was taken into account to get high photo-
The good chemical stability and special geometry of the TiO2 catalytic performance, since too low loading of MoS2 leads to
nanobelt makes it especially suitable for the preparation of a insufficient visible light absorption, while excessive loading of
double heterostructure. As mentioned previously, the surface MoS2 blocks the photo-electron transfer between the core part
coarsened TiO2 nanobelts easily form double heterostructures of the TiO2 nanobelt and the MoS2 shell. The unique advan-
because the post-formed TiO2 nanoparticles on the TiO2 nano- tages of 2D–1D surface heterostructure based on TiO2 nano-
belt act as new energetic sites for adsorption and reaction. For belts was further demonstrated the by rGO@TiO2 (Figure 11c)
example, the above mentioned Ag nanoparticles can be depos- and g-C3N4@TiO2 nanobelt surface heterostructures assembled
ited randomly onto the TiO2 nanoparticles on the TiO2 nanobelt through the photoreduction method.[177] The GO was photo-
surface to randomly form a double heterostructure. The double catalytically reduced and wrapped around the TiO2 nanobelt
heterostructure can further enhance the photocatalytic activity in situ to form the 2D–1D surface heterostructure. As a result,
of TiO2 nanobelts in a dramatic fashion. This effect has been the surface area and electrical conductivity were significantly
proven by the controllable synthesis of a new double surface enhanced, which ensures high photocatalytic performance.
heterostructure of Bi2O3/Bi4Ti3O12/TiO2 nanobelt, fabricated More interestingly, the graphene wrapped TiO2 nanobelt can be
through wet-chemical coprecipitation and post-solid reac- a potential functional platform for the construction of ternary
tion.[169] The Bi4Ti3O12 interphase plays the role of a bridge, heterostructures.[179]
linking Bi2O3 and TiO2. Under light irradiation, the special Conducting Polymer–TiO2 Nanobelt Heterostructures: Con-
crystalline structure of the Bi4Ti3O12 interphase can effectively ducting polymers are at the center of investigation for a wide
promote the separation of photogenerated electrons and holes. range of applications. Their unique applications and extensive
The second oxide phase in the TiO2 nanobelt surface heter- interest are mainly due to the elevated conductivity, catalytic
ostructure can also be in the form of 1D nanostructure, such activity, gas sensitivity and optoelectronic properties. These
as ZnO nanorod@TiO2 nanobelt surface heterostructure and conducting polymers typically include polyaniline (PANi),
branched TiO2 nanobelt.[170] These 1D–1D surface heterostruc- polypyrrole (PPy),poly(3-hexylthiophene) (P3TH), and so on.[180]
tures further prove that TiO2 nanobelts are the ideal platforms When the conducting polymers are coupled with semiconduc-
for the construction of various surface heterostructures. tors, the hybrid heterostructures formed have some unique
Chalcogenide–TiO2 Nanobelt Surface Heterostructures: Dif- physical and electrical properties. The large conjugated systems
ferent from metal oxide semiconductors, chalcogenides are in these polymer molecules are especially good for charge-
typically narrow bandgap semiconductors which have wide and carrier mobility, and this effect can be further enhanced when
strong absorption from UV- to visible-light regions.[171] They are assembled with 1D nanostructures, demonstrated to be a very
also famous for the controllability of their bandgap and band appealing strategy.
structure.[172] Chalcogenides have been widely used for many However, the synthesis of one-dimensional core–shell struc-
years as photocatalysts in water splitting and the degradation tures with high surface area and effective interfacial charge
of organic pollutants.[173] By constructing chalcogenide nano- transfer is still challenging. Hydrogen titanate nanobelts have
structures on the surface of TiO2 nanobelts, the light harvesting a hydrophilic surface, and aniline, which is the precursor
behavior and photocatalytic performance of TiO2 nanobelts can monomer of PANi, is water soluble. Therefore, the p-type PANi
be obviously increased.[174] Under light irradiation, electrons in can form a uniform shell on the surface of hydrogen titanate
the valence band of chalcogenides are excited and easily trans- nanobelts through in situ polymerization.[181] Individual
ferred into TiO2 nanobelts. TiO2 nanobelts with good charge hydrogen titanate–PANi core-shell nanowires are a prototypical
mobility have the ability to guide electron transfer. For example, system, for which electrical properties have been explored.
PbS quantum dot-decorated (QD-decorated) TiO2 nanobelts The highly thermal stable hydrogen titanate–PANi hetero-
exhibited extended light absorption up to a wavelength of structures have a wide light absorption range of 200–800 nm,
1400 nm (Figure 11a).[175] Meanwhile, the band structure of the dramatically larger than pure hydrogen titanate nanobelts
surface heterostructure can be tuned by adjusting the size of that have a light absorption cut-off edge at about 400 nm. Its
the PbS quantum dots. This quantum dot surface heterostruc- resistance strongly depends on temperature. When the temper-
ture displayed an increased fluorescence lifetime and charge- ature is higher than 100 °C, the heterostructure shows ohmic
transfer rate constant. contact properties, with high resistance. Up to a temperature

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 19
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

Figure 11. a) TEM images demonstrate the configuration of the PbS quantum dots/TiO2 nanobelt heterostructure. Reproduced with permission.[175]
Copyright 2010, American Chemical Society. b) Low-magnification TEM images of TiO2@MoS2 heterostructures (50 wt% of MoS2) and the corre-
sponding SAED pattern. Reproduced with permission.[176] c) AFM scan showing the topography of the graphene oxide (GO) sheets deposited on mica,
and TEM and HRTEM images showing the morphology of surface-coarsened TiO2 nanobelt and rGO/TiO2 nanobelts. (lower-right) Dashed lines in
HRTEM image outline a portion of the interface between the TiO2 nanobelt and TiO2 nanoparticle, as well as that between TiO2 nanoparticle and rGO
sheet; (inset) schematic model of the rGO/TiO2 nanobelt heterostructure. Reproduced with permission.[177a]

of 100 °C, the resistance of the heterostructure decreases dra- is more favorable for electron delocalization. At temperatures
matically with increases in the temperature. The reason is that higher than 100 °C, the water molecules are released and the
at temperatures lower than 100 °C, the PANi displays a more chain alignment of the PANi polymer is disturbed.
p-type nature and the hole concentration is increased by dona- Although still very few reports exist on polymer–TiO2
tion from the PANi molecule toward the water molecules. Mean- nanobelt heterostructures, with the progress in synthesis
while, the heating process-induced molecular conformation of TiO2 nanobelts, and great demand of organic–inorganic

20 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
Figure 12. a) Representative SEM and TEM images of Bi2WO6/TiO2 nanobelt heterostructures. Photocatalytic degradation of MO in the presence of
P25, TiO2 nanobelts, Bi2WO6 nanosheets and Bi2WO6/TiO2 nanobelt heterostructures under b) UV light, c) visible light, d) near-infrared light and
e) simulated solar light. f) The experimental visible and near-infrared absorption spectrum of TiO2 nanobelts, Bi2WO6 nanosheets and Bi2WO6/TiO2
nanobelt heterostructures. g) Visible and near-infrared absorption spectra of TiO2 nanobelts, Bi2WO6 nanosheets, and Bi2WO6/TiO2 nanobelt hetero-
structures. a–g) Reproduced with permission.[183]

one-dimensional composites, polymer–TiO2 nanobelt hetero- 6.1. Bi2WO6 Nanosheet–TiO2 Nanobelt Surface
structures will attract much attention. Heterostructures

The above mentioned challenge may be solved with the

6. Full Solar Light Spectrum Photocatalysts Based recently-reported broad spectrum photocatalyst consisting of
hybrid Bi2WO6 nanosheets–TiO2 nanobelts surface heterostruc-
on TiO2 Nanobelts
tures. This hybrid heterostructure was demonstrated to pos-
Solar energy — free, clean, and inexhaustible — is the most sess the ability to harvest UV, visible, and near-infrared light
promising renewable energy source for human beings.[182] to decompose organic contaminants in aqueous solution. The
The enhancement of solar energy harvesting and conver- hydrothermally synthesized Bi2WO6 nanosheets showed good
sion is a major goal in photo-electric conversion fields. This visible light absorption and even had a strong absorption in
challenging mission is especially important in photocatalysis the near-infrared light range, indicating excellent visible and
and photovoltaic devices because of the highly increased near-infrared light catalytic performance (Figure 12e).[183] The
environmental problems and energy crisis. Thus, the search proportion of MO degraded using Bi2WO6 nanosheets–TiO2
for photocatalysts that can harvest light across the full solar nanobelt surface heterostructures under UV, visible, and near-
spectrum, from UV to near-infrared wavelengths, and effi- infrared light irradiation for 120 min was up to 24%, 84% and
ciently convert the solar energy for further applications is 58%, respectively. The photocatalytic activity using visible light
critical. may come from the Bi2WO6 itself,[85,184] while the near infrared

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 21
DOI: 10.1002/adma.201405589
 [Link]
[Link]

derivatives and ferrocyanide, as well as the reduction of Ag+,

REVIEW

photocatalytic ability is suggested to originate from its oxygen

vacancies. Cu+ and oxygen under visible light and NIR light (<860 nm).
Achieving the use of solar light to the utmost extent is the One main and important feature of gold and silver is the local-
main direction and object of photocatalysis. However, full spec- ized surface plasmon-induced (LSP–induced) characteristic
trum photocatalysts are difficult to realize because of the lack bands of optical attenuation at visible light and infrared wave-
of infrared-light active photocatalysts. Therefore, the discovery lengths. Nishijima and co-workers reported photoelectric con-
of the near-infrared photocatalytic activity of Bi2WO6 should version from light in the visible to NIR range using electrodes
provoke new investigations into to highly efficient utilization of in which Au nanorods were elaborately arrayed on the surface
solar energy. of a TiO2 single crystal.[141] This system was thought to be able
to use light with wavelengths from 800 to 1200 nm. Further-
more, the LSP significantly increased the photocurrent genera-
6.2. Carbon Quantum Dot–TiO2 Nanobelt Surface tion. Qin et al. have reported a novel NIR photocatalyst of Yb3+
Heterostructures and Tm3+ co-doped YF3 nanocrystal core/TiO2 shell core/shell
nanoparticles.[189] The core/shell nanoparticles could effectively
Carbon quantum dots (CQDs) with typical sizes below 10 nm decompose methylene blue under the irradiation of 980 nm
are newly emerging carbon nanomaterials with promising laser, which was ascribed to the upconversion mechanism.
applications because of their benign, abundant and inexpensive Actually, upconversion should be an effective way to use NIR
nature. The structure of CQDs, i.e., amorphous or crystalline, light energy, because materials with upconversion ability can
depends on the synthetic methods.[185] Notably, CQDs exhibit be easily realized by doping of materials like NaYF4.[124,125,138a]
both up- and down-conversion PL (photoluminescence), and However, such upconverting materials also have their intrinsic
photogenerated electron transfer ability. The broad NIR absorp- disadvantages, such as the narrow wavelength range for NIR
tion spectrum, upconversion PL property, and laser excitation light excitation.
make CQDs an excellent candidate photocatalyst for harvesting
NIR light.
The photocatalytic properties of CQDs under NIR light 7. Concluding Remarks
illumination were demonstrated recently by the highly selec-
tive oxidation of benzyl alcohol using CQDs as a photocata- The synthetic methods to produce TiO2 nanobelts are well
lysts under NIR light irradiation.[186] In this work, CQDs were developed, even at large-scales. This provides good support for
obtained from alkali-assisted electrochemical methods and the wide application of TiO2 nanobelts. Because of their specific
had sizes in the range 1–4 nm. The conversion efficiency and geometry, TiO2 nanobelts have special physical and chemical
selectivity of CQDs can reach values as high as 92% and 100%, properties which causes surprising performance in various
respectively, well beyond those of carbon nanoparticles (with fields such as environment protection, solar cell and water split-
sizes in the range 100–150 nm and conversion efficiency and ting for hydrogen fuels, and numerous sensors. However, TiO2
selectivity values of 71% and 78%, respectively). The photoelec- nanobelts by themselves have some drawbacks, which can be
trochemical results showed that, under NIR light irradiation, overcome using TiO2 nanobelt surface heterostructures. Such
the CQD electrode exhibited an obvious photocurrent response, heterostructures have been demonstrated to provide a very
while the carbon nanoparticles displayed no obvious photocur- effective way to solve the problems faced by researchers when
rent under the same conditions. CQDs displayed a broad lumi- using bare TiO2 nanobelts, e.g., lack of absorbance in the vis-
nescence peak at about 400 nm, with excitation wavelength of ible and IR ranges of the spectrum. While TiO2 nanobelt surface
830 nm, and its PL can be efficiently quenched by either elec- heterostructures have already become a well-accepted strategy to
tron-acceptor or electron-donor molecules in solution. All these improve performance, studies on these surface heterostructures
investigations confirm that CQDs can be an ideal near-infrared show further potential for development beyond the current state
light photocatalyst and can be used to modify TiO2 nanobelts of the art. Directions of future research should focus on the fol-
to form surface heterostructures. Our recent work regarding lowing aspects: 1) theoretical understanding of the mechanism
CQD–TiO2 nanobelt heterostructures proved that they possess of the interfacial bandgap tuning of TiO2 nanobelt surface heter-
UV–vis–NIR light photocatalytic activity, which opens a new ostructures, based on calculations and simulations, and reaching
door for the synthesis of full solar spectrum light photocatalyst an efficient charge-carrier separation based on a profound theo-
using carbon-based quantum dots.[187] retical understanding; 2) discovery, design and fabrication of vis-
ible and infrared light photocatalysts with high performance, to
realize the synthesis of high performance photocatalysts that are
6.3. TiO2 Nanobelt Surface Heterostructures based on Other active in the whole solar light spectrum; 3) design and assembly
Materials of devices based on the 1D nanostructured TiO2 nanobelt sur-
face heterostructure for practical applications.
Materials with NIR or infrared photocatalytic properties are
not only limited to the above Bi2WO6 and CQDs. Gold clus-
ters also exhibit photocatalytic performance under NIR light
irradiation, and gold cluster-modified TiO2 shows both vis- Acknowledgements
ible light and NIR light activity.[188] Au25 cluster-modified TiO2 Z.Z. and J.T. contributed equally to this work. The authors are thankful
nanoparticles can efficiently drive the oxidation of phenol for funding from the National Natural Science Foundation of China

22 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

(Grant No. 51372142), the Innovation Research Group (IRG: 51321091) [29] A. L. Linsebigler, G. Lu, J. T. Yates Jr., Chem. Rev. 1995, 95, 735.

REVIEW
and the “100 Talents Program” of the Chinese Academy of Sciences. [30] A. G. Thomas, K. L. Syres, Chem. Soc. Rev. 2012, 41, 4207.
The authors also acknowledge the support from the “Thousands Talents [31] W. J. Zhou, L. G. Gai, P. G. Hu, J. J. Cui, X. Y. Liu, D. Z. Wang,
Program” for pioneer researcher and his innovation team, China. A.C. G. H. Li, H. D. Jiang, D. Liu, H. Liu, J. Y. Wang, CrystEngComm
acknowledges the Spanish MINECO for financial support under contract 2011, 13, 6643.
ENE2013–46624-C4–3-R. [32] N. Fu, Y. Wu, Z. Jin, G. Lu, Langmuir 2010, 26, 447.
[33] W. J. Zhou, H. Liu, J. Y. Wang, D. Liu, G. J. Du, J. J. Cui, ACS Appl.
Received: December 6, 2014 Mater. Inter. 2010, 2, 2385.
Revised: February 9, 2015 [34] D. Li, H. Haneda, S. Hishita, N. Ohashi, Chem. Mater. 2005, 17,
Published online: 2596.
[35] Z. L. Wang, Adv. Mater. 2003, 15, 432.
[36] a) I. Kaplan-Ashiri, R. Tenne, J. Cluster Sci. 2007, 18, 549;
[1] M. Segal, Nat. Nanotechnol. 2009, 4, 612. b) B. Wen, J. E. Sader, J. J. Boland, Phys. Rev. Lett. 2008, 101,
[2] A. Fujishima, K. Honda, Nature 1972, 238, 37. 175502.
[3] a) M. R. Hoffmann, S. T. Martin, W. Y. Choi, D. W. Bahnemann, [37] a) H. Ni, X. Li, Nanotechnology 2006, 17, 3591; b) W. Mai,
Chem. Rev. 1995, 95, 69; b) A. L. Linsebigler, G. Q. Lu, J. T. Yates, Z. L. Wang, Appl Phys Lett 2006, 89, 073112.
Chem. Rev. 1995, 95, 735; c) K. Kalyanasundaram, M. Gratzel, [38] Y. Wang, R. Wang, C. Guo, J. Miao, Y. Tian, T. Ren, Q. Liu,
Coordin. Chem. Rev. 1998, 177, 347. Nanoscale 2012, 4, 1545.
[4] M. Grätzel, J. Photochem. Photobiol. C 2003, 4, 145. [39] M. Humar, D. Arčon, P. Umek, M. Škarabot, I. Muševič, G. Bregar,
[5] J. Gong, Y. Li, Z. Hu, Z. Zhou, Y. Deng, J. Phys. Chem. C 2010, 114, Nanotechnology 2006, 17, 3869.
9970. [40] P. Umek, P. Cevc, A. Jesih, A. Gloter, C. P. Ewels, D. Arcčon, Chem.
[6] a) K. S. Mun, S. D. Alvarez, W. Y. Choi, M. J. Sailor, ACS Nano Mater. 2005, 17, 5945.
2010, 4, 2070; b) S. J. Bao, C. M. Li, J. F. Zang, X. Q. Cui, Y. Qiao, [41] X. Wu, S. S. Amin, T. T. Xu, J. Mater. Res. 2011, 25, 935.
J. Guo, Adv. Funct. Mater. 2008, 18, 591. [42] M. Chang, J. R. Deka, C. H. Lin, C. C. Chung, Key Eng. Mater. 2010,
[7] J. Zhang, M. J. Li, Z. C. Feng, J. Chen, C. Li, J. Phys. Chem. B 2006, 437, 66.
110, 927. [43] G. He, J. Eckert, W. Löser, M. Hagiwara, Solid State Commun.
[8] a) T. L. Thompson, J. T. Yates Jr., Top. Catal. 2005, 35, 197; 2004, 129, 711.
b) D. Aspnes, Surf. Sci. 1983, 132, 406; c) A. G. Agrios, P. Pichat, J. [44] A. Selloni, Nat. Mater. 2008, 7, 613.
Photochem. Photobiol. A 2006, 180, 130. [45] H. G. Yang, G. Liu, S. Z. Qiao, C. H. Sun, Y. G. Jin, S. C. Smith,
[9] J. Hu, T. W. Odom, C. M. Lieber, Acc. Chem. Res. 1999, 32, 435. J. Zou, H. M. Cheng, G. Q. Lu, J. Am. Chem. Soc. 2009, 131,
[10] Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim, 4078.
H. Yan, Adv. Mater. 2003, 15, 353. [46] M. Liu, L. Piao, L. Zhao, S. Ju, Z. Yan, T. He, C. Zhou, W. Wang,
[11] Z. W. Pan, Z. R. Dai, Z. L. Wang, Science 2001, 291, 1947. Chem. Commun. 2010, 46, 1664.
[12] J. Tian, Z. Zhao, A. Kumar, R. I. Boughton, H. Liu, Chem. Soc. Rev. [47] M. A. Fox, C. C. Chen, J. Am. Chem. Soc. 1981, 103, 6757.
2014, 43, 6920. [48] N. G. Petrik, Z. Zhang, Y. Du, Z. Dohnálek, I. Lyubinetsky,
[13] M. Ni, M. K. Leung, D. Y. Leung, K. Sumathy, Renewable Sustain- G. A. Kimmel, J. Phys. Chem. C 2009, 113, 12407.
able Energy Rev. 2007, 11, 401. [49] T. L. Thompson, J. T. Yates, Chem. Rev. 2006, 106, 4428.
[14] Z. Y. Wu, G. Ouvrard, P. Gressier, C. R. Natoli, Phys. Rev. B 1997, [50] T. D. Nguyen-Phan, M. B. Song, E. J. Kim, E. W. Shin, Res. Chem.
55, 10382. Intermed. 2010, 36, 613.
[15] C. Howard, T. Sabine, F. Dickson, Acta Crystallogr. B 1991, 47, 462. [51] a) J. M. Macak, H. Tsuchiya, P. Schmuki, Angew. Chem. Int. Ed.
[16] W. J. Zhou, X. Y. Liu, J. J. Cui, D. Liu, J. Li, H. D. Jiang, J. Y. Wang, 2005, 44, 2100; b) J. Wang, L. Zhao, V. S.-Y. Lin, Z. Lin, J. Mater.
H. Liu, CrystEngComm 2011, 13, 4557. Chem. 2009, 19, 3682.
[17] Y. M. Wang, G. J. Du, H. Liu, D. Liu, S. B. Qin, N. Wang, C. G. Hu, [52] N. T. Q. Hoa, E.-T. Kim, Electrochem. Solid-State Lett. 2008, 11,
X. T. Tao, J. Jiao, J. Y. Wang, Z. L. Wang, Adv. Funct. Mater. 2008, K1.
18, 1131. [53] a) Y. W. L. Lim, Y. X. Tang, Y. H. Cheng, Z. Chen, Nanoscale 2010,
[18] N. Q. Wu, J. Wang, D. Tafen, H. Wang, J. G. Zheng, J. P. Lewis, 2, 2751; b) D. R. Zhang, C. W. Kim, Y. S. Kang, J. Phys. Chem. C
X. G. Liu, S. S. Leonard, A. Manivannan, J. Am. Chem. Soc. 2010, 2010, 114, 8294; c) H. G. Yang, H. C. Zeng, J. Am. Chem. Soc.
132, 6679. 2005, 127, 270; d) L. X. Pang, X. Y. Wang, X. D. Tang, Solid State
[19] W. E. Kaden, T. Wu, W. A. Kunkel, S. L. Anderson, Science 2009, Sci. 2015, 39, 29.
326, 826. [54] D. Wu, J. Liu, X. N. Zhao, A. D. Li, Y. F. Chen, N. B. Ming, Chem.
[20] R. Asahi, Y. Taga, W. Mannstadt, A. Freeman, Phys. Rev. B 2000, 61, Mater. 2006, 18, 547.
7459. [55] F. Pan, K. Wu, H. X. Li, G. Q. Xu, W. Chen, Chem. Eur. J. 2014, 20,
[21] H. Peng, J. Li, J. Phys. Chem. C 2008, 112, 20241. 15095.
[22] T. He, F. Pan, Z. Xi, X. Zhang, H. Zhang, Z. Wang, M. Zhao, [56] S. I. Cha, K. H. Hwang, Y. H. Kim, M. J. Yun, S. H. Seo, Y. J. Shin,
S. Yan, Y. Xia, J. Phys. Chem. C 2010, 114, 9234. J. H. Moon, D. Y. Lee, Nanoscale 2013, 5, 753.
[23] W. J. Zhou, H. Liu, R. I. Boughton, G. J. Du, J. J. Lin, J. Y. Wang, [57] K. Kiatkittipong, C. Ye, J. Scott, R. Amal, Cryst. Growth Des. 2010,
D. Liu, J. Mater. Chem. 2010, 20, 5993. 10, 3618.
[24] T. He, M. Zhao, X. Zhang, H. Zhang, Z. Wang, Z. Xi, X. Liu, S. Yan, [58] R. C. Xie, J. K. Shang, J. Mater. Sci. 2007, 42, 6583.
Y. Xia, L. Mei, J. Phys. Chem. C 2009, 113, 13610. [59] J. Y. Liao, B. X. Lei, H. Y. Chen, D. B. Kuang, C. Y. Su, Energy
[25] M. A. Fox, M. T. Dulay, Chem. Rev. 1993, 93, 341. Environ. Sci. 2012, 5, 5750.
[26] J. Bisquert, V. S. Vikhrenko, J. Phys. Chem. B 2004, 108, 2313. [60] Q. H. Chen, H. L. Liu, Y. J. Xin, X. W. Cheng, J. Zhang, J. J. Li,
[27] J. Chen, B. Li, J. Zheng, S. Jia, J. Zhao, H. Jing, Z. Zhu, J. Phys. P. Wang, H. J. Li, Electrochim. Acta 2013, 99, 152.
Chem. C 2011, 115, 7104. [61] A. Mohammadpour, P. R. Waghmare, S. K. Mitra, K. Shankar, ACS
[28] X. Chen, S. S. Mao, Chem. Rev. 2007, 107, 2891. Nano 2010, 4, 7421.

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 23
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

[62] J. Ge, Y. Li, Adv. Funct. Mater. 2004, 14, 157. [88] S. Y. Choi, M. Mamak, G. von Freymann, N. Chopra, G. A. Ozin,
[63] a) J. M. Wu, H. C. Shih, W. T. Wu, Chem. Phys. Lett. 2005, 413, 490; Nano Lett. 2006, 6, 2456.
b) J. M. Wu, H. C. Shih, W. T. Wu, Nanotechnology 2006, 17, 105; [89] J. Wang, Chem. Rev. 2008, 108, 814.
c) C. Xiong, J. Kenneth, J. Balkus, Chem. Mater. 2005, 17, 5136. [90] J. J. Cui, D. H. Sun, W. J. Zhou, H. Liu, P. G. Hu, N. Ren,
[64] S. S. Amin, A. W. Nicholls, T. T. Xu, Nanotechnology 2007, 18, H. M. Qin, Z. Huang, J. J. Lin, H. Y. Ma, Phys. Chem. Chem. Phys.
445609. 2011, 13, 9232.
[65] J. Shi, X. Wang, Cryst. Growth Des. 2011, 11, 949. [91] J. J. Cui, D. H. Sun, S. W. Chen, W. J. Zhou, P. G. Hu, H. Liu,
[66] D. Xia, Y. B. Jiang, X. He, S. R. J. Brueck, Appl. Phys. Lett. 2010, 97, Z. Huang, J. Mater. Chem. 2011, 21, 10633.
223106. [92] J. J. Cui, Y. K. Ge, S. W. Chen, H. S. Zhao, H. Liu, Z. Huang,
[67] S. S. Kim, C. Chun, J. C. Hong, D. Y. Kim, J. Mater. Chem. 2006, H. D. Jiang, J. Chen, J. Mater. Chem. B 2013, 1, 2072.
16, 370. [93] a) S. R. Morrison, Sens. Actuators 1982, 2, 329; b) A. Azad,
[68] a) K. Hashimoto, H. Irie, A. Fujishima, Jpn. J. Appl. Phys. 2005, 44, S. Akbar, S. Mhaisalkar, L. Birkefeld, K. Goto, J. Electrochem. Soc.
8269; b) Z. Xie, X. X. Liu, W. P. Wang, X. J. Wang, C. Liu, Q. Xie, 1992, 139, 3690.
Z. C. Li, Z. J. Zhang, Nanoenergy 2015, 11, 400; c) D. Zhong, [94] G. Eranna, B. Joshi, D. Runthala, R. Gupta, Crit. Rev. Solid State
B. Cai, X. L. Wang, Z. Yang, Y. D. Xing, S. Miao, W. H. Zhang, C. Li, Mater. Sci. 2004, 29, 111.
Nanoenergy 2015, 11, 409. [95] S. R. Morrison, Sens. Actuators 1987, 12, 425.
[69] X. Yu, X. Han, Z. H. Zhao, J. Zhang, W. B. Guo, C. F. Pan, A. X. Li, [96] H. F. Lu, F. Li, G. Liu, Z. G. Chen, D. W. Wang, H. T. Fang,
H. Liu, Z. L. Wang, Nanoenergy 2015, 11, 19. G. Q. Lu, Z. H. Jiang, H. M. Cheng, Nanotechnology 2008, 19, 405504.
[70] a) J. Zhao, T. Wu, K. Wu, K. Oikawa, H. Hidaka, N. Serpone, [97] G. Chaudhari, D. Bambole, A. Bodade, P. Padole, J. Mater. Sci.
Environ. Sci. Technol. 1998, 32, 2394; b) Y. Liang, H. Wang, 2006, 41, 4860.
H. S. Casalongue, Z. Chen, H. Dai, Nano Res. 2010, 3, 701. [98] S. Yoriya, H. E. Prakasam, O. K. Varghese, K. Shankar, M. Paulose,
[71] H. Lachheb, E. Puzenat, A. Houas, M. Ksibi, E. Elaloui, C. Guillard, G. K. Mor, T. J. Latempa, C. A. Grimes, Sens. Lett. 2006, 4, 334.
J. M. Herrmann, Appl. Catal. B 2002, 39, 75. [99] H.-M. Lin, C.-H. Keng, C.-Y. Tung, Nanostruct. Mater. 1997, 9, 747.
[72] a) F. Li, X. Li, M. Hou, K. Cheah, W. Choy, Appl. Catal. A 2005, 285, [100] A. M. Ruiz, G. Sakai, A. Cornet, K. Shimanoe, J. R. Morante,
181; b) H. Parham, F. Khoshnam, J. Chem. Technol. Biotechnol. N. Yamazoe, Sens. Actuators B 2003, 93, 509.
2013, 88, 1736. [101] a) E. Traversa, M. L. Di Vona, S. Licoccia, M. Sacerdoti,
[73] Y. Ding, C. Yang, L. Zhu, J. Zhang, J. Hazard. Mater. 2010, 175, 96. M. C. Carotta, L. Crema, G. Martinelli, J. Sol–Gel Sci. Technol. 2001,
[74] C. F. Lin, C. H. Wu, Z. N. Onn, J. Hazard. Mater. 2008, 154, 1033. 22, 167; b) I. D. Kim, A. Rothschild, B. H. Lee, D. Y. Kim, S. M. Jo,
[75] S. Fukahori, H. Ichiura, T. Kitaoka, H. Tanaka, Environ. Sci. Technol. H. L. Tuller, Nano Lett. 2006, 6, 2009.
2003, 37, 1048. [102] J. Huang, Q. Wan, Sensors 2009, 9, 9903.
[76] W. J. Zhou, G. J. Du, P. G. Hu, Y. Q. Yin, J. H. Li, J. H. Yu, [103] P. Q. Hu, G. J. Du, W. J. Zhou, J. J. Cui, J. J. Lin, H. Liu, D. Liu,
G. C. Wang, J. X. Wang, H. Liu, J. Y. Wang, H. Zhang, J. Hazard. J. Y. Wang, S. W. Chen, ACS Appl. Mater. Interfaces 2010, 2, 3263.
Mater. 2011, 197, 19. [104] T. J. Hsueh, C. L. Hsu, S. J. Chang, I. Chen, Sens. Actuators B 2007,
[77] a) J. Remillard, J. McBride, K. Nietering, A. Drews, X. Zhang, 126, 473.
J. Phys. Chem. B 2000, 104, 4440; b) T. Hirakawa, Y. Nosaka, Lang- [105] L. J. Bie, X. N. Yan, J. Yin, Y. Q. Duan, Z. H. Yuan, Sens. Actuators
muir 2002, 18, 3247. B 2007, 126, 604.
[78] Q. H. Chen, H. L. Liu, Y. J. Xin, X. W. Cheng, J. J. Li, Appl. Surf. Sci. [106] Y. Chen, L. Nie, X. Xue, Y. Wang, T. Wang, Appl. Phys. Lett. 2006,
2013, 264, 476. 88, 083105.
[79] P. K. Dubey, R. S. Tiwari, O. N. Srivastava, A. S. K. Sinha, [107] L. Francioso, A. Taurino, A. Forleo, P. Siciliano, Sens. Actuators B
Int. J. Nanosci. 2011, 10, 723. 2008, 130, 70.
[80] a) U. Bach, D. Lupo, P. Comte, J. Moser, F. Weissörtel, J. Salbeck, [108] a) C. Xiang, M. Li, M. Zhi, A. Manivannan, N. Wu, J. Mater. Chem.
H. Spreitzer, M. Grätzel, Nature 1998, 395, 583; b) N.-G. Park, 2012, 22, 19161; b) Y. Luo, D. Kong, J. Luo, S. Chen, D. Zhang,
J. Van de Lagemaat, A. Frank, J. Phys. Chem. B 2000, 104, 8989; K. Qiu, X. Qi, H. Zhang, C. M. Li, T. Yu, RSC Adv. 2013, 3, 14413.
c) X. Zhang, F. Liu, Q. L. Huang, G. Zhou, Z. S. Wang, J. Phys. [109] a) H. Huang, J. Fang, Y. Xia, X. Tao, Y. Gan, J. Du, W. Zhu,
Chem. C 2011, 115, 12665; d) B. O’Regan, M. Grfitzeli, Nature W. Zhang, J. Mater. Chem. A 2013, 1, 2495; b) H. Ming, Y. Yan,
1991, 353, 737. J. Ming, X. Li, Q. Zhou, H. Huang, J. Zheng, RSC Adv. 2014, 4,
[81] K. D. Benkstein, N. Kopidakis, J. Van de Lagemaat, A. Frank, 12971; c) W. Q. Han, X. L. Wang, Appl. Phys. Lett. 2010, 97, 243104;
J. Phys. Chem. B 2003, 107, 7759. d) B. Liu, D. Deng, J. Y. Lee, E. S. Aydil, J. Mater. Res. 2010, 25,
[82] a) M. Adachi, Y. Murata, J. Takao, J. Jiu, M. Sakamoto, F. Wang, J. 1588; e) Y. Luo, J. Luo, W. Zhou, X. Qi, H. Zhang, D. Y. W. Yu,
Am. Chem. Soc. 2004, 126, 14943; b) D. Kuang, J. Brillet, P. Chen, C. M. Li, H. J. Fan, T. Yu, J. Mater. Chem. A 2013, 1, 273; f) M. Fehse,
M. Takata, S. Uchida, H. Miura, K. Sumioka, S. M. Zakeeruddin, S. Cavaliere, P. Lippens, I. Savych, A. Iadecola, L. Monconduit,
M. Grätzel, ACS Nano 2008, 2, 1113; c) K. Zhu, N. R. Neale, D. Jones, J. Rozière, F. Fischer, C. Tessier, J. Phys. Chem. C 2013,
A. Miedaner, A. J. Frank, Nano Lett. 2007, 7, 69. 117, 13827; g) P. Gao, D. Bao, Y. Wang, Y. Chen, L. Wang, S. Yang,
[83] J. Jiu, S. Isoda, F. Wang, M. Adachi, J. Phys. Chem. B 2006, 110, G. Chen, G. Li, Y. Sun, W. Qin, ACS Appl. Mater. Interfaces 2013,
2087. 5, 368; h) Z. Yang, G. Du, Q. Meng, Z. Guo, X. Yu, Z. Chen,
[84] X. Feng, K. Zhu, A. J. Frank, C. A. Grimes, T. E. Mallouk, Angew. T. Guo, R. Zeng, J. Mater. Chem. 2012, 22, 5848; i) Y. Qiao, X. Hu,
Chem. Int. Ed. 2012, 51, 2727. Y. Huang, J. Nanopart. Res. 2012, 14, 1.
[85] K. Pan, Q. Zhang, Q. Wang, Z. Liu, D. Wang, J. Li, Y. Bai, Thin Solid [110] D. Deng, M. G. Kim, J. Y. Lee, J. Cho, Energy Environ. Sci. 2009, 2,
Films 2007, 515, 4085. 818.
[86] a) S. Liu, A. Chen, Langmuir 2005, 21, 8409; b) E. Topoglidis, [111] T. Beuvier, M. Richard-Plouet, M. Mancini-Le Granvalet,
C. J. Campbell, A. E. Cass, J. R. Durrant, Langmuir 2001, 17, 7899. T. Brousse, O. Crosnier, L. Brohan, Inorg. Chem. 2010, 49, 8457.
[87] a) J. I. Hahm, C. M. Lieber, Nano Lett. 2004, 4, 51; b) Z. Gao, [112] M. Mao, L. Mei, D. Guo, L. Wu, D. Zhang, Q. Li, T. Wang,
A. Agarwal, A. D. Trigg, N. Singh, C. Fang, C. H. Tung, Y. Fan, Nanoscale 2014, 6, 12350.
K. D. Buddharaju, J. Kong, Anal. Chem. 2007, 79, 3291. [113] H. Wang, C. Guan, X. Wang, H. J. Fan, Small 2014.

24 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589
 [Link]
[Link]

REVIEW
[114] a) T. H. Ji, L. L. Li, H. Liang, J. Y. Sun, J. W. Cai, G. L. Ge, Rare Met. [146] J. Li, S. K. Cushing, P. Zheng, T. Senty, F. Meng, A. D. Bristow,
Mater. Eng. 2009, 38, 1030; b) T. H. Ji, F. Yang, Y. Y. Lv, J. Y. Zhou, A. Manivannan, N. Q. Wu, J. Am. Chem. Soc. 2014, 136,
J. Y. Sun, Mater. Lett. 2009, 63, 2044; c) X. Feng, K. Shankar, 8438.
M. Paulose, C. A. Grimes, Angew. Chem. Int. Ed. 2009, 121, 8239; [147] S. K. Cushing, J. Li, F. Meng, T. R. Senty, S. Suri, M. Zhi, M. Li,
d) S. Jeon, P. V. Braun, Chem. Mater. 2003, 15, 1256. A. D. Bristow, N. Wu, J. Am. Chem. Soc. 2012, 134, 15033.
[115] J. C. S. Wu, C. H. Chen, J. Photochem. Photobiol. A 2004, 163, 509. [148] J. Li, S. K. Cushing, J. Bright, F. Meng, T. R. Senty, P. Zheng,
[116] F. Fabregat-Santiago, E. M. Barea, J. Bisquert, G. K. Mor, A. D. Bristow, N. Q. Wu, ACS Catalysis 2013, 3, 47.
K. Shankar, C. A. Grimes, J. Am. Chem. Soc. 2008, 130, 11312. [149] J. Tersoff, Phys. Rev. Lett. 1984, 52, 465.
[117] S. Klosek, D. Raftery, J. Phys. Chem. B 2001, 105, 2815. [150] H. Chen, S. Chen, X. Quan, Y. Zhang, Environ. Sci. Technol. 2009,
[118] L. F. Cui, F. Huang, M. T. Niu, L. W. Zeng, J. Xu, Y. S. Wang, J. Mol. 44, 451.
Catal. A 2010, 326, 1. [151] C.-T. Sah, R. N. Noyce, W. Shockley, Proc. IRE 1957, 45, 1228.
[119] S. V. Chong, J. Xia, N. Suresh, K. Yamaki, K. Kadowaki, Solid State [152] a) V. Subramanian, E. E. Wolf, P. V. Kamat, J. Am. Chem. Soc. 2004,
Commun. 2008, 148, 345. 126, 4943; b) D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-
[120] H. Zhu, J. Tao, X. Dong, J. Phys. Chem. C 2010, 114, 2873. Espriu, A. F. Carley, A. A. Herzing, M. Watanabe, C. J. Kiely,
[121] T. H. Ji, Y. Liu, H. Zhao, H. Y. Du, J. Y. Sun, G. L. Ge, J. Solid State D. W. Knight, G. J. Hutchings, Science 2006, 311, 362.
Chem. 2010, 183, 584. [153] C. Y. Su, L. Liu, M. Y. Zhang, Y. Zhang, C. L. Shao, CrystEngComm
[122] R. Janisch, P. Gopal, N. A. Spaldin, J. Phys.: Condens. Matter 2005, 2012, 14, 3989.
17, R657. [154] a) Y. L. Liu, L. Zhong, Z. Y. Peng, Y. Cai, Y. B. Song, W. Chen, Cryst-
[123] W. Choi, A. Termin, M. R. Hoffmann, J. Phys. Chem. C 1994, 98, EngComm 2011, 13, 5467; b) T. Burgi, A. Baiker, J. Phys. Chem. B
13669. 2002, 106, 10649.
[124] H. Y. Zhang, T. H. Ji, Y. F. Liu, J. W. Cai, J. Phys. Chem. C 2008, 112, [155] D. Z. Wang, W. J. Zhou, P. G. Hu, Y. Guan, L. M. Chen, J. H. Li,
8604. G. C. Wang, H. Liu, J. Y. Wang, G. Z. Cao, H. D. Jiang, J. Colloid
[125] a) H. B. Li, H. Zhao, Y. Sheng, J. Huang, Y. H. Song, K. Y. Zheng, Interface Sci. 2012, 388, 144.
Q. S. Huo, H. F. Zou, J. Nanopart. Res. 2012, 14; b) T. H. Ji, [156] a) R. Ma, T. Sasaki, Y. Bando, J. Am. Chem. Soc. 2004, 126, 10382;
F. Yang, H. Y. Du, H. Guo, J. S. Yang, J. Rare Earths 2010, 28, 529. b) T. Akita, M. Okumura, K. Tanaka, K. Ohkuma, M. Kohyama,
[126] F. Wang, Y. Han, C. S. Lim, Y. Lu, J. Wang, J. Xu, H. Chen, T. Koyanagi, M. Date, S. Tsubota, M. Haruta, Surf. Interface Anal.
C. Zhang, M. Hong, X. Liu, Nature 2010, 463, 1061. 2005, 37, 265; c) C. H. Lin, J. H. Chao, C. H. Liu, J. C. Chang,
[127] F. Wang, X. Liu, Chem. Soc. Rev. 2009, 38, 976. F. C. Wang, Langmuir 2008, 24, 9907; d) F. C. Wang, C. H. Liu,
[128] Z. L. He, W. X. Que, J. Chen, X. T. Yin, Y. C. He, J. B. Ren, ACS C. W. Liu, J. H. Chao, C. H. Lin, J. Phys. Chem. C 2009, 113, 13832.
Appl. Mater. Interfaces 2012, 4, 6815. [157] a) E. Formo, E. Lee, D. Campbell, Y. N. Xia, Nano Lett. 2008, 8,
[129] Q. Xiao, L. L. Ouyang, L. Gao, C. Yao, Appl. Surf. Sci. 2011, 257, 668; b) H. Y. He, P. Yang, C. C. Jia, Y. P. Miao, J. Zhao, Y. Y. Du,
3652. Funct. Mater. Lett. 2014, 7, 1450064.
[130] J. Wang, D. N. Tafen, J. P. Lewis, Z. L. Hong, A. Manivannan, [158] Y. Wang, G. Du, H. Liu, D. Liu, S. Qin, J. Wang, X. Tao, M. Jiang,
M. J. Zhi, M. Li, N. Q. Wu, J. Am. Chem. Soc. 2009, 131, 12290. Z. L. Wang, J. Nanosci. Nanotechnol. 2009, 9, 2119.
[131] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 2001, [159] W. Zhou, D. Wang, H. Liu, J. Wang, Sci. Adv. Mater. 2014, 6,
293, 269. 538.
[132] a) O. Diwald, T. L. Thompson, T. Zubkov, E. G. Goralski, [160] T. T. Jiang, C. C. Jia, L. C. Zhang, S. R. He, Y. H. Sang, H. D. Li,
S. D. Walck, J. T. Yates, J. Phys. Chem. B 2004, 108, 6004; Y. Q. Li, X. H. Xu, H. Liu, Nanoscale 2015, 7, 209.
b) F. Dong, W. Zhao, Z. Wu, Nanotechnology 2008, 19, 365607; [161] a) L. Huang, F. Peng, H. J. Wang, H. Yu, Z. Li, Catal. Commun.
c) H. Wang, J. Lewis, J. Phys.: Condens. Matter 2006, 18, 421; 2009, 10, 1839; b) L. F. Cui, F. Huang, M. T. Niu, L. W. Zeng, J. Xu,
d) M. Sathish, B. Viswanathan, R. Viswanath, C. S. Gopinath, Y. S. Wang, J. Mol. Catal. A 2010, 326, 1; c) W. J. Zhou, H. Liu,
Chem. Mater. 2005, 17, 6349. J. Y. Wang, D. Liu, G. J. Du, J. J. Cui, ACS Appl. Mater. Interfaces
[133] N. Serpone, J. Phys. Chem. B 2006, 110, 24287. 2010, 2, 2385; d) L. Zhu, G. C. Liu, X. C. Duan, Z. J. Zhang,
[134] Z. Y. Han, J. C. Zhang, W. L. Cao, Mater. Lett. 2012, 84, 34. J. Mater. Chem. 2010, 25, 1278.
[135] D. N. Tafen, J. Wang, N. Q. Wu, J. P. Lewis, Appl. Phys. Lett. 2009, [162] a) C. H. Wang, C. L. Shao, X. T. Zhang, Y. C. Liu, Inorg. Chem. 2009,
94. 48, 7261; b) J. B. Mu, B. Chen, M. Y. Zhang, Z. C. Guo, P. Zhang,
[136] W. J. Zhou, G. J. Du, P. G. Hu, G. H. Li, D. Z. Wang, H. Liu, Z. Y. Zhang, Y. Y. Sun, C. L. Shao, Y. C. Liu, ACS Appl. Mater. Inter-
J. Y. Wang, R. I. Boughton, D. Liu, H. D. Jiang, J. Mater. Chem. faces 2012, 4, 424; c) L. P. Wu, X. J. Li, Z. H. Yuan, Y. Chen, Catal.
2011, 21, 7937. Commun. 2009, 11, 67; d) W. Zeng, T. M. Liu, Z. C. Wang, J. Mater.
[137] A. Fujishima, T. N. Rao, D. A. Tryk, J. Photochem. Photobiol. C Chem. 2012, 22, 3544.
2000, 1, 1. [163] L. X. Yang, S. L. Luo, Y. Li, Y. Xiao, Q. Kang, Q. Y. Cai, Environ. Sci.
[138] M. R. Hoffmann, S. T. Martin, W. Choi, D. W. Bahnemann, Chem. Technol. 2010, 44, 7641.
Rev. 1995, 95, 69. [164] J. J. Lin, J. X. Shen, R. J. Wang, J. J. Cui, W. J. Zhou, P. G. Hu,
[139] M.-F. Joubert, Opt. Mater. 1999, 11, 181. D. O. Liu, H. Liu, J. Y. Wang, R. I. Boughton, Y. Z. Yue, J. Mater.
[140] a) F. Wang, X. Liu, J. Am. Chem. Soc. 2008, 130, 5642; Chem. 2011, 21, 5106.
b) J. C. Boyer, F. Vetrone, L. A. Cuccia, J. A. Capobianco, J. Am. [165] J. Tian, Y. Sang, Z. Zhao, W. Zhou, D. Wang, X. Kang, H. Liu,
Chem. Soc. 2006, 128, 7444. J. Wang, S. Chen, H. Cai, H. Huang, Small 2013, 9, 3864.
[141] Y. Nishijima, K. Ueno, Y. Yokota, K. Murakoshi, H. Misawa, J. Phys. [166] R. Liu, P. Hu, S. Chen, Appl. Surf. Sci. 2012, 258, 9805.
Chem. Lett. 2010, 1, 2031. [167] J. Zhang, W. X. Liu, X. W. Wang, X. Q. Wang, B. Hu, H. Liu,
[142] T. Hirakawa, P. V. Kamat, Langmuir 2004, 20, 5645. Appl. Surf. Sci. 2013, 282, 84.
[143] J. Gao, X. Ren, D. Chen, F. Tang, J. Ren, Scr. Mater. 2007, 57, [168] J. J. Cui, S. W. Chen, H. Liu, Z. Huang, J. Solid State Electrochem.
687. 2014, 18, 2693.
[144] S. K. Cushing, N. Wu, Interface 2013, 22, 41. [169] Z. H. Zhao, J. Tian, D. Z. Wang, X. L. Kang, Y. H. Sang, H. Liu,
[145] J. Li, S. K. Cushing, P. Zheng, F. Meng, D. Chu, N. Q. Wu, Nat. J. Y. Wang, S. W. Chen, R. I. Boughton, H. D. Jiang, J. Mater. Chem.
Commun. 2013, 4, 2651. 2012, 22, 23395.

Adv. Mater. 2015, © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [Link] 25
DOI: 10.1002/adma.201405589
 [Link]
[Link]
REVIEW

[170] a) Z. H. Zhao, D. Z. Wang, X. L Kang, Y. H. Sang, H. Liu, Energy [179] J. Y. Zhang, F. X. Xiao, G. C. Xiao, B. Liu, Nanoscale 2014, 6,
Environ. Focus 2014, 3, 404; b) D. Sarkar, K. K. Chattopadhyay, ACS 11293.
Appl. Mater. Interfaces 2014, 6, 10044. [180] a) A. Malinauskas, J. Malinauskiene, A. Ramanavičius, Nanotech-
[171] G. Wertheim, F. DiSalvo, D. Buchanan, Solid State Commun. 1973, nology 2005, 16, R51; b) H. D. Tran, D. Li, R. B. Kaner, Adv. Mater.
13, 1225. 2009, 21, 1487.
[172] S. Chen, X. Gong, C. G. Duan, Z. Q. Zhu, J. H. Chu, A. Walsh, [181] L. G. Gai, G. J. Du, Z. Y. Zuo, Y. M. Wang, D. Liu, H. Liu, J. Phys.
Y. G. Yao, J. Ma, S. H. Wei, Phys. Rev. B 2011, 83, 245202. Chem. C 2009, 113, 7610.
[173] N. Zheng, X. Bu, H. Vu, P. Feng, Angew. Chem. 2005, 117, 5433. [182] D. Bahnemann, Sol. Energy 2004, 77, 445.
[174] a) S. V. Chong, N. Suresh, J. Xia, N. Al-Salim, H. Idriss, J. Phys. [183] J. Tian, Y. Sang, G. Yu, H. Jiang, X. Mu, H. Liu, Adv. Mater. 2013,
Chem. C 2007, 111, 10389; b) H. Li, M. Eastman, R. Schaller, 25, 5075.
W. Hudson, J. Jiao, J. Nanosci. Nanotechnol. 2011, 11, 8517; [184] a) H. Fu, C. Pan, W. Yao, Y. Zhu, J. Phys. Chem. B 2005, 109, 22432;
c) J. Y. Zhang, F. X. Xiao, G. C. Xiao, B. Liu, New J. Chem. 2015. b) C. Zhang, Y. Zhu, Chem. Mater. 2005, 17, 3537.
[175] D. F. Wang, H. G. Zhao, N. Q. Wu, M. A. El Khakani, D. L. Ma, [185] a) S. N. Baker, G. A. Baker, Angew. Chem. Int. Ed. 2010, 49, 6726;
J. Phys. Chem. Lett. 2010, 1, 1030. b) W. Zhang, H. Zhu, S. Yu, H. Yang, Adv. Mater. 2012, 24, 2263.
[176] W. J. Zhou, Z. Y. Yin, Y. P. Du, X. Huang, Z. Y. Zeng, Z. X. Fan, [186] H. Li, R. Liu, S. Lian, Y. Liu, H. Huang, Z. Kang, Nanoscale 2013,
H. Liu, J. Y. Wang, H. Zhang, Small 2013, 9, 140. 5, 3289.
[177] a) Y. Sang, Z. Zhao, J. Tian, P. Hao, H. Jiang, H. Liu, J. P. Claverie, [187] J. Tian, Y. H. Leng, Z. H. Zhao, Y. Xia, Y. H. Sang, P. Hao, J. Zhan,
Small 2014, 10, 3775; b) X. Zhong, M. M. Jin, H. Q. Dong, L. Liu, M. C. Li, H. Liu, Nanoenergy 2015, 11, 419.
L. Wang, H. Y. Yu, S. Leng, G. L. Zhuang, X. N. Li, J. G. Wang, J. [188] A. Kogo, N. Sakai, T. Tatsuma, Electrochem. Commun. 2010, 12,
Solid State Chem. 2014, 220, 54. 996.
[178] F. Meng, J. Li, S. K. Cushing, M. Zhi, N. Wu, J. Am. Chem. Soc. [189] W. Qin, D. Zhang, D. Zhao, L. Wang, K. Zheng, Chem. Commun.
2013, 135, 10286. 2010, 46, 2304.

26 [Link] © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2015,
DOI: 10.1002/adma.201405589