Hierarchical ZnO aggregates assembled by orderly aligned nanorods for dye-sensitized solar cells

Article abstract of DOI:10.1039/c2ce26460a

Hierarchical ZnO aggregates assembled by orderly aligned nanorods were prepared via a facile solvothermal method using diethylene glycol as solvent and zinc acetate dehydrate as precursor. Time dependent trails evidenced that the formation of the hierarchical aggregates experienced a multistep self-assembly process. Moreover, it was found that the reaction medium plays an important role in the assembling process and the diameter of the product could be tuned by simply altering the precursor dosage. The hierarchical product was further applied as scattering layer in bi-layered dye-sensitized solar cell (DSSC), and a high conversion efficiency of 5.2% was demonstrated, indicating a substantial improvement compared with the nanoparticle cell of 2.7%. Based on the optical and electrochemical investigations, the high conversion efficiency was mainly ascribed to the unique hierarchical structure of the ZnO aggregates. The rough surface of the nanorod subunits could enhance the dye loading capacity and the aggregates with micrometer sized diameter could improve the scattering effect. Moreover, the orderly aligned nanorods could minimize the grain boundaries, suppress the surface recombination and provide a direct pathway for fast electron transport, which therefore enhance the collection efficiency of the photoelectrons.

Full text of DOI:10.1039/c2ce26460a

CrystEngComm
View Article Online
View Journal | View Issue
PAPER
Hierarchical ZnO aggregates assembled by orderly
aligned nanorods for dye-sensitized solar cells3
Cite this: CrystEngComm, 2013, 15,
1210
Dapeng Wu,ab Zhiyong Gao,ab Fang Xu,a Jiuli Chang,a Wenguang Tao,a Jinjin He,a
Shuyan Gaoa and Kai Jiang*ab
Hierarchical ZnO aggregates assembled by orderly aligned nanorods were prepared via a facile
solvothermal method using diethylene glycol as solvent and zinc acetate dehydrate as precursor. Time
dependent trails evidenced that the formation of the hierarchical aggregates experienced a multistep self-
assembly process. Moreover, it was found that the reaction medium plays an important role in the
assembling process and the diameter of the product could be tuned by simply altering the precursor
dosage. The hierarchical product was further applied as scattering layer in bi-layered dye-sensitized solar
cell (DSSC), and a high conversion efficiency of 5.2% was demonstrated, indicating a substantial
improvement compared with the nanoparticle cell of 2.7%. Based on the optical and electrochemical
investigations, the high conversion efficiency was mainly ascribed to the unique hierarchical structure of
the ZnO aggregates. The rough surface of the nanorod subunits could enhance the dye loading capacity
and the aggregates with micrometer sized diameter could improve the scattering effect. Moreover, the
orderly aligned nanorods could minimize the grain boundaries, suppress the surface recombination and
provide a direct pathway for fast electron transport, which therefore enhance the collection efficiency of
the photoelectrons.
Received 8th September 2012,
Accepted 18th November 2012
DOI: 10.1039/c2ce26460a
www.rsc.org/crystengcomm
cross-linking function of poly(sodium 4-styrenesulfonate) and
ZnO hollow microspheres were also prepared via a biopolymer
1. Introduction
Zinc oxide, which has a direct band gap of y3.3 eV and large
exciton binding energy of 60 meV, is one of the most versatile
functional semiconductors. ZnO nanostructures with particu-
larly engineered size and morphology could bring forth
desirable physical and chemical properties. Especially, ZnO
hierarchical structures with various assembling patterns are of
great research interest because these complicated structures
could ensure their prominent performances in different
functional devices.1–4 However, different from the hierarchical
structural assembled from zero dimensional nanoparticles,
the great mass and volume of the primary building blocks
make them difficult to assemble into orderly aligned hier-
archical structures with the assistant of weak interactions such
as the van der Waals and electrostatic forces.5–7 Generally, the
assembling of the subunits is assisted with the macroscale
forces from organic molecules, for example, ZnO microspheres
assembled by the nanorods were fabricated with the aid of the
directed crystal growth and mediated self-assembly of well
aligned nanorods.8,9
Dye-sensitized solar cell (DSSC) represents a new genera-
tion solar energy conversion device which have attracted much
research attention in the past two decades due to their
appealing qualities such as low producing cost, short energy
payback time and convenience for multiple application
purposes.10,11 ZnO was regarded as the best alternative for
TiO2 materials because of the similar electron band structure,
excellent bulk electron mobility and sophisticated structural/
morphological engineering strategies.12–17 Therefore, these
unique advantages of ZnO were expected to expand the
possibilities in term of designing the DSSC photoanode. To
date, a vast variety of ZnO nanostructures such as nanopar-
ticles, tetrapods, nanosheets and hierarchical aggregates were
designed and fabricated for the photoanode materials.18–27
However, the highest PCE of ZnO based DSSCs merely reached
to y7.5% adopting polydisperse submicron sized hierarchical
ZnO aggregates as photoanode and N719 as sensitizer, which
is still far less than the PCE of the TiO2 based DSSCs.28
In order to achieve high PCE on ZnO based DSSCs, many
schemes were developed to enhance the light harvesting and
the charge collection efficiency. There are generally two
methods to improve the harvesting of the incident light:
mixing the nanoparticles film with submicron sized ZnO
aSchool of Chemistry and Chemical Engineering, Henan Key Laboratory of
Photovoltaic Materials, Xinxiang, Henan, P. R. China
bEngineering Technology Research Center of Motive Power and Key Materials of
Henan Province, Xinxiang, Henan 453007, P. R. China.
E-mail: jiangkai6898@126.com; Fax: +86 373 3326209
Electronic supplementary information (ESI) available: XRD characterizations of
3
the intermediate samples; SEM images of the Film-S2. See DOI: 10.1039/
c2ce26460a
1210 | CrystEngComm, 2013, 15, 1210–1217
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
reaction time was set at 6 h and the other reaction parameters
were kept unchanged.
particles as scattering centers, and depositing additional
scattering layer on the nanoparticles film.29–32 Nevertheless,
the larger size particles inevitably reduce the overall surface
area and the dye loading capacity of the film. To mediate this
competition, ZnO submicron sized aggregates consisting of
10–20 nm nanocrystals were employed as dual-functional
material to improve the scattering of the incident light without
sacrificing the accessible surface area.33,34 To achieve high
collection efficiency, one dimensional ZnO nanostructures are
intentionally introduced to reduce the recombination rate by
offering direct pathway for electron transport.35,36 However,
these one dimensional ZnO structures usually possess limited
surface area which consequently results in low short circuit
current density.37,38 Therefore, it is of great significance to
integrate the required qualities of large surface area, promi-
nent scattering effect and high electron transport rate into the
ZnO based photoanode. In addition, due to the intrinsic
instability in acid ambient, the ZnO nanoparticles are
vulnerable to the ruthenium based dyes. Moreover, the dyes
containing complexing agents can also detach the Zn2+ from
the host lattice to generate Zn2+-dye agglomerations at the
interface, resulting in low injection efficiency and high
recombination rate.39–41 The modification of another protect-
ing layer (like SiO2, Al2O3, TiO2 and ZnS) could indeed enhance
the stability of ZnO surface and suppress the recombina-
tion.42–46 However it would inevitably bring forth injection
problem and extra fabrication cost.47 Therefore, there are still
many challenges to promote the PCE of ZnO based DSSCs.
This paper reported a facile fabrication of ZnO hierarchical
comprised of orderly aligned nanorods, and an interesting
self-assembly process was observed by the time dependent
trials. Moreover, the diameter of the products could be turned
by simply altering the precursor dosage. We intentionally
deposited the as-prepared hierarchical ZnO aggregates, which
are assembled by orderly aligned nanorods and have different
diameter ranges, on the nanoparticle film to construct the bi-
layered DSSCs. Due to the promoted scattering effect and the
suppressed recombination, a high PCE of 5.2% was demon-
strated in the bi-layered cell, which indicated a substantial
PCE enhancement compared with the cell derived from
monolayer ZnO nanoparticle (2.7%).
2.2. Preparation of 20–30 nm ZnO nanoparticles
0.5 g Zn(Ac)2?2H2O was dispersed into 40 mL absolute ethanol.
After stirring for 1 h, the suspension was transferred into 50
mL Teflon-lined autoclave and heated at 160 uC for 4 h. The
white powder was rinsed with distilled water and absolute
ethanol, and finally dried in vacuum for later use.
2.3. Material characterizations
The products were characterized by field-emission scanning
electron microscopy (FESEM, HITACHI, S4800), transmission
electron microscopy (TEM, FEI Tecnai F30) equipped with a
X-ray energy diffusion spectra with detection limit of 0.3–1%,
X-ray powder diffraction (XRD, Rigaku D/max-2500 diffract-
ometer with Cu Ka radiation, l = 0.1542 nm, 40 kV, 100 mA).
2.4. Fabrication of the DSSCs
Hydroxypropyl cellulose (Aldrich) was firstly added to diethy-
lene glycol with a concentration of y10 wt% to prepare the
paste. To prepare the nanoparticle slurry, y0.4 g the as-
prepared paste was added into 0.2 g ZnO nanoparticles, and
stirred vigorously to generate the homogeneous slurry. To
prepare the scattering layer slurries, 25 wt% nanoparticles
were mixed in the ZnO aggregates powders. Then, the slurries
were prepared in a similar process with the nanoparticle
slurry. The bi-layered films were constructed by the doctor-
blade method through two-step calcination. The nanoparticle
slurry was spread onto FTO glass substrate (TEC-8, LOF) with
adhesive tape to control the film thickness. After drying in air,
the film was heated up to 450 uC and maintained for 30 min.
After calcination, the ZnO aggregates slurries were deposited
on the nanoparticle layer and annealed with the same heating
profile. After cooled to y70 uC, the films were sensitized in a
ethanol solution of cis-bis(isothiocyanato)bis(2,29-bipyridy1-
4,49-dicarboxylate)
ruthenium(II)bis-tetrabutylammonim
(N719, Solaronix SA, Switzerland, 3 6 1024 M) for 50 min.
The dyed films were then sandwiched together with platinized
FTO counter electrodes and the electrolyte was then injected
into the cell from the edges by capillarity action. The content
of the electrolyte is 0.05 M LiI, 0.03 M I2, 0.5 M tert-
butylpyridine, 0.1 M GuNCS and 1 M BMII in a mixture of
85% acetonitrile and 15% valeronitrile.
2.5. Photovoltaic measurements
2. Experimental section
The photocurrent–voltage (I–V) measurements were performed
on Keithley 4200 semiconductor characterization system using
simulated AM 1.5 sunlight with an output power of 100 mW
cm22 produced by a solar simulator (Newport 69911). The
incident monochromatic photo-to-electron conversion effi-
ciency (IPCE) was recorded on a Keithley 2000 sourcemeter
under the irradiation of a 150 W tungsten lamp with 1/4 m
monochromator (Spectral Product DK240). The optical diffuse-
reflection spectra were measured by spectrophotometer
(HITACHI U-4100). The electrochemical impedance spectra
(EIS) were measured in dark and at an applied negative bias
voltage. The spectra were scanned at a frequency ranging from
100 kHz to 100 MHz with a computer controlled potentiostat
2.1. Preparation of the ZnO hierarchical aggregates
1 g zinc acetate dehydrate (Zn(Ac)2?2H2O) was dispersed into
40 mL diethylene glycol (Aldrich, 99%). After stirring at room
temperature for 1 h, the suspension was transferred into 50
mL Teflon-lined autoclave and heated at 160 uC for 6 h. The
residues were thoroughly rinsed with distilled water and
absolute ethanol by centrifugation. The final products were
dried in vacuum for later use. For the time dependent trails,
the intermediate samples were respectively collected at 0.5 h, 1
h, 2 h and 4 h to monitor the growing process. Moreover, the
Zn(Ac)2?2H2O dosage was also adjusted from 0.5 g to 1.5 g to
control the size and morphology of the products, in which the
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 1210–1217 | 1211

View Article Online
Paper
CrystEngComm
(EG&G, M283) equipped with a frequency response detector
(EG&G, FRD100).
3. Results and discussion
3.1. Morphology and structure characterizations
Fig. 1a displays the as-prepared hierarchical ZnO aggregates
which are assembled by several individual microspheres and
possess a wide diameter distribution of 4–8 mm. Moreover,
some individual microspheres could be also observed with
smaller diameter of 1–3 mm. As depicted in Fig. 1b, a broken
aggregate reveals the hierarchical structures are assembled by
orderly aligned nanorods about 500 nm to 1 mm in length. In
addition, the cross section (c area) and the surface (d area) of
the aggregates were further magnified and shown in Fig. 1c
and d. Fig. 1c shows the orderly aligned nanorods are radially
assembled and have rough surface. In addition, Fig. 1d
discloses the nanorods are about 10–30 nm in diameter and
with typical hexagonal edges. Fig. 2 shows that the XRD
diffraction peaks of the aggregates match well with the
standard values (JCPDS No. 36-1451) and no impurity is
detected, suggesting the product has high purity and wurtzite
structure. Moreover, it is interesting to find that the full width
of the (002) peak at the half-maximum is the narrowest when
compared with the other diffraction peaks, inferring that the
as-generated ZnO subunit was promoted along the [0001]
direction to form the rod-like structures.48
Fig. 2 XRD pattern of the as-prepared hierarchical ZnO aggregates.
Fig. 3b shows the lattice fringes are clear and the interplanar
spacing is measured as 0.26 nm, indicating the ZnO nanorod
is growing along the [0001] direction. Moreover, Fig. 3c
indicates the surface of the nanorod is not smooth but
distributed with holes about several nanometers in diameter
(marked with white circles), which could improve the surface
roughness and the dye loading capacity. The corresponding
fast Fourier transfer (FFT) image (inset of Fig. 3a) reveals a dot
like pattern, indicating the nanorod is virtually single crystal
and stretching along the [0001] direction. Fig. 3d displays the
EDS characterization of the nanorod bundle, in spite of the C
and Cu peaks deriving from the copper grid, no impurity
elements are detected, confirming the sample has high purity.
Fig. 3a shows a nanorod bundle detached from the
hierarchical ZnO aggregate. The nanorods have a rough
surface and are radially aligned to form a fan-shaped bundle.
Moreover, the bright and narrow slits (marked in white
squares) indicate there are channels existing among the rods,
which may facilitate the diffusion of the N719 solution during
the sensitizing. Fig. 3b and c respectively presents the HRTEM
image of the b and c area marked with black squares in Fig. 3a.
3.2. Growth mechanism and size control
In order to reveal the growth mechanism of the hierarchical
ZnO aggregates, time dependent trails were carried out to
monitor the grow process. After 0.5 h reaction, Fig. 4a shows
the firstly generated precipitate was composed of nanoparti-
cles and about several nanometers in diameter. As the reaction
reached to 1 h, ZnO nanorods about 200–300 nm in length
were formed (Fig. 4b). Moreover, at this stage, ZnO nanopar-
ticles about 10–20 nm in diameter were found coexisting with
Fig. 1 SEM images of the as-prepared hierarchical ZnO aggregates with (a) low,
Fig. 3 (a) TEM image, (b, c) HRTEM images and (d) EDS characterization of the
(b) intermediate and (c, d) high magnifications.
nanorod bundle detached from the aggregates. The inset of (a) is the FFT image.
1212 | CrystEngComm, 2013, 15, 1210–1217
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
charged (000-1) facet of another particle is apt to link together
by the electrostatic attraction. Therefore, the finally formed
nanorods all stretching along the [0001] direction by means of
the oriented attachment.50–52 As the reaction preceded, the
nanoparticles would further fuse together to generate high
quality single crystal nanorods. Considering nanoparticles are
the primary units for the later assembly, it is reasonable to
understand that the nanorods possess rough surface.
Although the reason why the orderly aligned nanorod bundles
all taking radial assembling pattern remains a mystery, the
driving force may also attribute to the minimization of the
system free energy (Step 2).53 Subsequently, considering the
nanorods are radially aligned to develop the fanshaped
bundles, it is reasonable to anticipate that the later assembled
products should exhibit spherical shapes (Step 3). As depicted
in Scheme 1, the individual growth of the spherical product
leads to the formation of individual microspheres. Meanwhile,
the aggregation growth results in large size aggregates
containing several microspheres. The aggregation growth
may further improve the connectivity among the individual
microspheres and lead to a better electron transport.
In addition, as displayed in Fig. 5, the nonpolar diethylene
glycol medium is found to exert great importance for the self-
assembly process. In order to eliminate the effect of the crystal
water contained in the precursor, the Zn(Ac)2?2H2O was firstly
dehydrated under 100 uC for 2 h before transferred into the
solvothermal reaction system. After 6 h, the diameter of the as-
obtained aggregates is generally reduced to less than 2 mm
(Fig. 5a). The inset shows the aggregate comprises of
nanoparticles about 10–20 nm in diameter, which is different
from nanorod subunits with hexagonal cross-section. As 0.5
mL water was introduced to the reaction, although the
morphology of the product is slightly changed, many small
particles are formed beside the aggregates (Fig. 5b). If the
water dosage was further elevated to 1 mL, the assembling
process is obviously jeopardized and only several individual
aggregates survive. Moreover, the diameter of the subunit
increased to 30–100 nm as shown in the inset (Fig. 5c). The
assembling was totally destroyed and no aggregate was formed
Fig. 4 SEM images of the intermediate products collected at different times: (a)
0.5 h, (b) 1 h, (c) 2 h and (d) 4 h.
the nanorods (inset of Fig. 4b). When the reaction proceeded
to 2 h, nanorod bundles about 500 nm in length were
observed, and the inset image indicates the bundles possess
rough surface and assembled by nanoparticles (Fig. 4c). After 4
h, micrometer sized hierarchical aggregates appeared in the
sample (Fig. 4d). Meanwhile, the inset image suggests many
nanorod bundles are littering aside the hierarchical aggre-
gates, indicating the bundles were not fully assembled yet. The
corresponding XRD investigation (Fig. S1, ESI ) shows that the
3
nanoparticles obtained at the early stage are wurtzite ZnO
having poor crystal quality. In addition, the crystal quality of
the intermediate ZnO products is gradually improved as
increasing the reaction time.
As illustrated in Scheme 1, a multistep self-assembly
mechanism is proposed to elucidate the growth of the
hierarchical structures. At the initial stage, the decomposition
of Zn(Ac)2?2H2O generates the ZnO nucleus. The nonpolar
diethylene glycol medium strongly inhibits the anisotropic
growth of the ZnO crystals.49 Therefore, ZnO nanoparticles are
firstly formed in the reaction, and the continuous decompos-
ing of the precursor leads to the growth of the particle size.
Because of the (0001) facet of the ZnO nanoparticle,
terminated by Zn atoms, is positively charged, the negatively
Fig. 5 SEM images of the products prepared with different water/diethylene
glycol ratios (volume/volume (mL)): (a) 0 : 40 (the Zn(Ac)2?2H2O was dehy-
drated under 100 uC for 2 h prior to the solvothermal reaction), (b) 0.5 : 39.5, (c)
1 : 39, (d) 2 : 38, (e) 10 : 30 and (f) 40 : 0. The insets represent the magnified
images.
Scheme 1 Illustration of the multistep self-assembly growth mechanism from
ZnO nanoparticles to the final hierarchical aggregates.
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 1210–1217 | 1213

View Article Online
Paper
CrystEngComm
3.3. Photovoltaic performances
In order to investigate the photovoltaic performances of the as-
prepared hierarchical ZnO aggregates, we compared four cells
with different film structures. The cell derived from a
monolayer of ZnO nanoparticles (y16 mm) is labeled as Cell-
NP. The hierarchical ZnO samples prepared with Zn(Ac)2?2H2O
dosage of 0.5, 1 and 1.5 g are respectively deposited as scatter
layers (y10 mm) on ZnO nanoparticle films (y6 mm) to
construct the bi-layered cells and denoted as Cell-S1, Cell-S2
and Cell-S3. The photoanode structures of Cell-S2 are
displayed in Fig. S2, ESI. The cross section view depicted in
3
Fig. S2a, ESI, shows the underlayer is closely packed with
3
Fig. 6 SEM images of the hierarchical ZnO aggregates prepared with different
Zn(Ac)2?2H2O dosages of (a–c) 0.5 g and (d–f) 1.5 g.
nanoparticles and the overlayer is composed of both large size
hierarchical aggregates and nanoparticles which are inten-
tionally mixed to improve the connectivity among the
individual aggregates. Fig. S2b, ESI, shows the surface of
3
the scattering layer, it is found that the gaps among the
aggregates are filled with nanoparticles to allow better
when 2 mL water was introduced to the reaction system
(Fig. 5d). Moreover, if the water dosage was further increased,
micrometer sized ZnO rods with sharp tips and hexagonal
cross-sections were formed, indicating the anisotropic growth
was recovered in the solvent with high polarity (Fig. 5e and f).
Therefore, in spite of the function to suppress the anisotropic
growth of the ZnO crystal, the diethylene glycol reaction
system could also stabilize the intermediate products (nano-
particles, nanorods and the rod bundles) and prevent them
from irregularly aggregating probably due to its high viscosity,
which leads to the highly ordered assembling of the subunits
and finally shape the hierarchical ZnO aggregates composing
of orderly aligned nanorods.
connectivity. Fig. S2c, ESI, proves the nanoparticles in the
3
underlayer are generally 20–30 nm in diameter. Moreover, the
dye loading capacity of the films was evaluated by desorbing
the dyed films in 0.5 M NaOH aqueous solution (Table 1). The
capacity of Cell-NP is determined as 2.4 6 1027 mol cm22
,
while Cell-S1 and Cell-S2 have the similar capacity of 2.1 and
2.0 6 1027 mol cm22, respectively. The nanorod subunits of
the sample prepared with 1.5 g Zn(Ac)2?2H2O dosage possess
large diameter and smooth surface which results in low
surface area. Therefore, Cell-S3 exhibits the lowest dye loading
capacity of 1.7 6 1027 mol cm22
.
Fig. 7 shows the photocurrent–voltage (I–V) curves of the
DSSCs and the detailed photovoltaic characteristics are
displayed in Table 1. Considering the open circuit voltage
(Voc) varies little among the four cells, the PCE variation is
mainly attributed to the change in the filling factor (FF) and
the short circuit current (Jsc). Conspicuously, the bi-layered
DSSCs give much higher FF than that of the Cell-NP, which
leads to prominent PCE improvements compared with the
monolayer nanoparticle cell. Typically, Cell-S2 demonstrates a
Jsc of 13.5 mA cm22, Voc of 614 mV, and FF of 62.7%. As a
result, the PCE reaches up to 5.2%, indicating a significant
improvement compared with Cell-NP (2.7%).
Fig. 8 depicts the UV-Vis reflection spectra of the single layer
films composed of ZnO nanoparticles and hierarchical ZnO
aggregates prepared under different precursor dosages (mixed
with 25 wt% nanoparticles). With similar film thickness of
y10 mm, the nanoparticle film exhibits high transparency,
especially at the region of 600–800 nm. On the other hand, the
films derived from the hierarchical aggregates all possess high
reflectance, which could improve the scattering of the incident
Fig. 6 a–c represent the SEM images of the sample prepared
with 0.5 g Zn(Ac)2?2H2O dosage. Fig. 6a shows that except for
several large size aggregates, the sample mainly consists of
individual microspheres about 2 mm in diameter. Fig. 6b
displays the cross section view of a broken microsphere,
suggesting the product is assembled by radially aligned
nanorods with very thin diameters. The surface view in
Fig. 6c indicates the orderly assembled nanorods are y10
nm in diameter. Fig. 6d–f display the product obtained with
1.5 g Zn(Ac)2?2H2O dosage. It is reasonable to understand that
with increasing precursor dosage, the sample mainly consists
of hierarchical aggregates containing several microspheres,
and the diameters are increased to 8–10 mm. The cross section
view displayed in Fig. 6e shows the orderly aligned nanorods
possess smooth surface. The surface view in Fig. 6f confirms
that the diameter of the nanorods increases to 30–60 nm.
Moreover, the perfect hexagonal cross sections of the
nanorods suggest the crystal quality is further improved.
Table 1 Photovoltaic performance of DSSCs (the active area for all the cells are about y0.16 cm2)
Cells
Jsc (mA cm22
)
Voc (mV)
FF (%)
PCE (%)
Dye absorption (mol cm22
)
Cell-NP
Cell-S1
Cell-S2
Cell-S3
8.9
11.7
13.5
12.6
614
586
614
596
49.4
63.0
62.7
61.6
2.7
4.3
5.2
4.6
2.4 6 1027
2.1 6 1027
2.0 6 1027
1.7 6 1027
1214 | CrystEngComm, 2013, 15, 1210–1217
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
Fig. 9 Incident photon-to-current conversion efficiency (IPCE) spectra for the as-
Fig. 7 I–V curves of the DSSCs measured under one sun illumination (AM 1.5G,
100 mW cm22).
prepared DSSCs.
the bi-layered DSSCs, which could be also ascribed to the
reduced injection and collection efficiency.
light. Moreover, based on previous reports, the scattering
effect of the particles is size dependent.54,55 The polydisperse
particles usually have better scattering effect compared with
the monodisperse ones because they could induce scattering
enhancement in a wider wavelength range.32
In order to interpret the electron recombination dynamics,
the cells were tested by electrochemical impedance spectra
(EIS). The measurements were performed in dark at different
applied negative bias voltage.58 Fig. 10a shows the Nyquist
plots of Cell-NP and Cell-S2 measured at an applied voltage of
20.6 V. Three typical semicircles appear from high to low
frequency are corresponded to the charge transfer at the
counter electrode, the electron transport and the recombina-
tion at the ZnO–dye–electrolyte interface and the Nernstian
diffusion in the electrolyte, respectively.59,60 The equivalent
circuit used to fit the series resistance of Rs, charge
recombination resistance of Rct and the chemical potential
of C is presented in Fig. 10a. The major concern of the current
research is the recombination (Rct2) occurring at the ZnO–dye–
electrolyte interface. The as-fitted data plots of Rct2 and C2 are
plotted in function of the applied negative bias voltage and
displayed in Fig. 10b and c. It is obvious that Cell-S2 has
Fig. 9 shows the incident photon-to-current conversion
efficiency (IPCE) spectra as a function of wavelength for the
DSSCs. The bi-layered cells all exhibit higher IPCE value.
Particularly, compared with Cell-NP, clear IPCE peaks are
observed in Cell-S1, Cell-S2, and Cell-S3 at the range from 500
nm to 700 nm, which derives from the improved scattering
effect. Moreover, considering the nanorods with rough surface
could increase the dye loading amount, Cell-S2 has the highest
absolute IPCE value in the entire wavelength region, which is
in agreement with its high Jsc. As mentioned above, the acid
dye solution could impact the surface integrity of the small
size ZnO particles, resulting in low injection rate and high
electron recombination.56,57 Therefore, although Cell-NP has
high dye loading amount, it shows the lowest IPCE value.
Similarly, Cell-S1 derives from the aggregates assembled by
thin nanorods which are also vulnerable to the acid dye
solution. Therefore, it exhibits a relative low IPCE value among
Fig. 10 Electrochemical impedance characterizations of Cell-NP and Cell-S2: (a)
the Nyquist plots of the EIS date measured at 20.60 V. (b) The charge
recombination resistance (Rct2), (c) chemical potential (C2) and (d) electron
recombination life time (tn) of the cells measured at different applied negative
bias voltages. The inset of (a) is the equivalent circuit.
Fig. 8 The UV-Vis reflection spectra of the single layer films composed of 20–30
nm ZnO nanoparticles, and the ZnO hierarchical aggregates prepared with
Zn(Ac)2?2H2O dosages of 0.5, 1 and 1.5 g.
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 1210–1217 | 1215

View Article Online
Paper
CrystEngComm
5 P. Hu, X. Zhang, N. Han, W. Xiang, Y. Cao and F. Yuan,
Cryst. Growth Des., 2011, 11, 1520.
higher Rct2 value than Cell-NP suggesting its advantages to
retard the charge recombinations. It is understandable that
the high quality and orderly aligned nanorods could not only
resist the surface etching in the acid dye but also provide a
direct pathway to facilitate the electron transport. Moreover,
Cell-S2 possesses higher C2 values, indicating it could
efficiently accumulate the charges and enhance the charge
density for a higher Jsc compared with that of Cell-NP.61,62 In
addition, the electron recombination life times (tn) could be
calculated by tn = Rct2 6 C2.63 As displayed in Fig. 10d, Cell-S2
has longer tn at the whole voltage range than Cell-NP. For
example, at the bias voltage roughly equals to the Voc of the
cells (20.6 V), Cell-S2 shows a tn of y14 ms which triples the
value of Cell-NP (y4 ms). The prolonged tn could allow more
effective electron transport and leads to a higher conversion
efficiency of Cell-S2.
6 T. Zhai, S. Xie, Y. Zhao, X. Sun, X. Lu, M. Yu, M. Xu, F. Xiao
and Y. Tong, CrystEngComm, 2012, 14, 1850.
7 S. Ran, Y. Zhu, H. Huang, B. Liang, J. Xu, B. Liu, J. Zhang,
Z. Xie, Z. Wang, J. Ye, D. Chen and G. Shen, CrystEngComm,
2012, 14, 3063.
8 M. Mo, J. C. Yu, L. Zhang and S. A. Li, Adv. Mater., 2005, 17,
756.
9 S. Gao, H. Zhang, X. Wang, R. Deng, D. Sun and G. Zheng,
J. Phys. Chem. B, 2006, 110, 15847.
10 B. Oregan and M. Gratzel, Nature, 1991, 353, 737.
11 M. Gratzel, Nature, 2001, 414, 338.
12 E. M. Kaidashew, M. Lorenz, H. von Wenckstern, A.
H. Rahm, C. Semmelhack, K. H. Han, G. Benndorf,
C. Bundesman, H. Hochmuth and M. Grundmann, Appl.
Phys. Lett., 2003, 82, 3901.
13 J. Halme, P. Vahermaa, K. Miettunen and P. Lund, Adv.
Mater., 2010, 22, E210.
14 C. Klingshirn, ChemPhysChem, 2007, 8, 782.
15 F. Xu and L. Sun, Energy Environ. Sci., 2011, 4, 818.
16 H. Zhang, R. Wu, Z. Chen, G. Liu, Z. Zhang and Z. Jiao,
CrystEngComm, 2012, 14, 1775.
17 L. Xu, Z. Li, Q. Cai, H. Wang, H. Gao, W. Lv and J. Liu,
CrystEngComm, 2010, 12, 2166.
18 H. Lu, S. Wang, L. Zhao, J. Li, B. Dong and Z. Xu, J. Mater.
Chem., 2011, 21, 4228.
19 M. Saito and S. Fujihara, Energy Environ. Sci., 2008, 1, 280.
20 W. Chen, Y. Qiu, Y. Zhong, K. S. Wong and S. Yang, J. Phys.
Chem. A, 2010, 114, 3127.
21 Z. Li, Y. Zhou, G. Xue, T. Yu, J. Liu and Z. Zou, J. Mater.
Chem., 2012, 22, 14341.
22 Y. Qiu, W. Chen and S. Yang, J. Mater. Chem., 2010, 20,
1001.
23 J. Lin, C. Lee and K. Ho, J. Mater. Chem., 2012, 22, 1270.
24 Q. Xie, J. Li, Q. Tian and R. Shi, J. Mater. Chem., 2012, 22,
13541.
25 Q. Zhang and G. Cao, J. Mater. Chem., 2011, 21, 6769.
26 Y. Sun, L. Wang, X. Yu and K. Chen, CrystEngComm, 2012,
14, 3199.
4. Conclusions
We have prepared hierarchical ZnO aggregates assembled by
orderly aligned nanorods. A multistep self-assembly mechan-
ism was proposed to elucidate the growth process of the
products. Moreover, the size of the product could be tuned by
simply altering the Zn(Ac)2?2H2O dosage. The aggregates were
further used as the scattering layer in DSSC applications. The
bi-layered DSSC demonstrated a high PCE of 5.2%, in contrast
to the monolayer nanoparticles cell of 2.7%. The PCE
enhancements could be attributed to (a) the nanorods with
rough surface could increase the dye loading capacity and (b)
the micrometer sized aggregates could enhance the scattering
effect, which improves the harvesting of the incident light.
Moreover, (c) the orderly aligned nanorods with high crystal
quality could suppress the surface recombination and offer
direct pathway for electron transport.
Acknowledgements
27 H. Chen, D. Kuang and C. Su, J. Mater. Chem., 2012, 22,
15475.
28 N. Memarian, I. Concina, A. Braga, S. M. Rozati, A. Vomiero
and G. Sberveglieri, Angew. Chem., Int. Ed., 2011, 50, 12321.
29 Y. Qiu, W. Chen and S. Yang, Angew. Chem., Int. Ed., 2010,
49, 4649.
30 Y. Z. Zheng, X. Tao, L. X. Wang, H. Xu, Q. Hou, W. L. Zhou
and J. F. Chen, Chem. Mater., 2010, 22, 928.
31 T. P. Chou, Q. F. Zhang, G. E. Fryxell and G. Z. Cao, Adv.
Mater., 2007, 19, 2588.
This work is supported by National Natural Science
Foundation of China (61204078 and 61176004), Basic and
Frontier Research Programs of Henan Provence (No.
092300410240 and No. 112300410094), Natural Science
Foundation of Education Department of Henan Province
(2010 B150012, 2011A150017).
32 Y. Zheng, J. Zhao, H. Zhang, J. Chen, W. Zhou and X. Tao,
Chem. Commun., 2011, 47, 11519.
References
1 X. D. Wang, J. H. Song, J. Liu and Z. L. Wang, Science, 2007,
316, 102.
33 Q. Zhang, T. P. Chou, B. Russo, S. A. Jenekhe and G. Z. Cao,
Angew. Chem., Int. Ed., 2008, 47, 2402.
2 M. H. Huang, S. Mao, H. Feick, H. Q. Yan, Y. Y. Wu,
H. Kind, E. Weber, R. Russo and P. D. Yang, Science, 2001,
292, 1897.
3 D. P. Lyvers, J. M. Moon, A. V. Kildishev, V. M. Shalaev and
A. Wei, ACS Nano, 2008, 2, 2569.
34 Q. Zhang, T. P. Chou, B. Russo, S. A. Jenekhe and G. Z. Cao,
Adv. Funct. Mater., 2008, 18, 1654.
35 Y. J. Lee, D. S. Ruby, D. W. Peters, B. B. McKenzie and J. W.
P. Hsu, Nano Lett., 2008, 8, 1501.
36 K. Wang, J. J. Chen, W. L. Zhou, Y. Zhang, Y. F. Yan, J. Pern
and A. Mascarenhas, Adv. Mater., 2008, 20, 3248.
37 C. Wu and J. Wu, J. Mater. Chem., 2011, 21, 13605.
4 H. M. Zhang, P. R. Liu, X. L. Liu, S. Q. Zhang, X. D. Yao, T.
C. An, R. Amal and H. J. Zhao, Langmuir, 2010, 26, 11226.
1216 | CrystEngComm, 2013, 15, 1210–1217
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
38 C. Wu, W. Liao and J. Wu, J. Mater. Chem., 2011, 21, 2871.
39 H. Horiuchi, R. Katoh, K. Hara, M. Yanagida, S. Murata,
H. Arakawa and M. Tachiya, J. Phys. Chem. B, 2003, 107,
2570.
40 N. A. Anderson, X. Ai and T. Q. Lian, J. Phys. Chem. B, 2003,
107, 14414.
51 L. Zhang, L. Yin, C. Wang, N. Lun and Y. Qi, ACS Appl.
Mater. Interfaces, 2010, 2, 1769.
52 C. Pacholski, A. Kornowski and H. Weller, Angew. Chem.,
Int. Ed., 2002, 41, 1188.
53 D. Zhang, L. Sun, J. Zhang, Z. Yan and C. H. Yan, Cryst.
Growth Des., 2008, 8, 3609.
54 D. Wu, F. Zhu, J. Li, H. Dong, Q. Li, K. Jiang and D. Xu, J.
Mater. Chem., 2012, 22, 11665.
55 I. Yu, Y. Kim, H. Kim, C. Lee and W. I. Lee, J. Mater. Chem.,
2011, 21, 532.
56 T. Chou, Q. Zhang and G. Cao, J. Phys. Chem. C, 2007, 111,
18804.
57 K. Keis, J. Lindgren, S. Lindquist and A. Hagfeldt,
Langmuir, 2000, 16, 4688.
58 K. Park, J. Xi, Q. Zhang and G. Cao, J. Phys. Chem. C, 2011,
115, 20992.
59 Q. Wang, S. Ito, M. Gratzel, F. Fabregat-Santiago, I. Mora-
Sero, J. Bisquert, T. Bessho and H. Imai, J. Phys. Chem. B,
2006, 110, 25210.
60 J. Bisquert, J. Phys. Chem. B, 2002, 106, 325.
61 L. De Marco, M. Manca, R. Buonsanti, R. Giannuzzi,
F. Malara, P. Pareo, L. Martiradonna, N. M. Giancaspro,
P. Cozzolibc and D. G. Gigli, J. Mater. Chem., 2011, 21,
13371.
62 X. Chen, T. Ling and X. Du, Nanoscale, 2012, 4, 5602.
63 J. Bisquert, A. Zaban, M. Greenshtein and I. Mora-Sero, J.
Am. Chem. Soc., 2004, 126, 13550.
41 J. B. Asbury, N. A. Anderson, E. C. Hao, X. Ai and T. Q. Lian,
J. Phys. Chem. B, 2003, 107, 7376.
42 Y. J. Shin, J. H. Lee, J. H. Park and N. G. Park, Chem. Lett.,
2007, 36, 1506.
43 V. Ganapathy, B. Karunagaran and S. Rhee, J. Power
Sources, 2010, 195, 5138.
44 D. Wu, Z. Gao, F. Xu, Z. Shi, W. Tao and K. Jiang,
CrystEngComm, 2012, 14, 7934.
45 M. Law, L. E. Greene, A. Radenovic, T. Kuykendall,
J. Liphardt and P. D. Yang, J. Phys. Chem. B, 2006, 110,
22652.
46 J. Qiu, M. Guo and X. Wang, ACS Appl. Mater. Interfaces,
2011, 3, 2358.
47 Q. Zhang, B. B. Garcia, X. Zhou, Y. H. Jeong and G. Cao,
Adv. Mater., 2010, 22, 2329.
48 U. Maiti, S. Maiti and K. Chattopadhyay, CrystEngComm,
2012, 14, 640.
49 D. Jezequel, J. Guenot, N. Jouini and F. Fievet, J. Mater. Res.,
1995, 10, 77.
50 Q. Zhang, S. Liu and S. Yu, J. Mater. Chem., 2009, 19, 191.
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 1210–1217 | 1217