RSC CP C4CP01200C 3....3 electrodes with a separator (NKK TF40, 35 mm) and polyvinyl alcohol (PVA)–LiCl gel as a solid electrolyte. A PVA–LiCl electrolyte was prepared by mixing

12214 | Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 This journal is© the Owner Societies 2014

Cite this:Phys.Chem.Chem.Phys.,

2014, 16, 12214

Construction of 3D V2O5/hydrogenated-WO3

nanotrees on tungsten foil for high-performancepseudocapacitors†

Fengmei Wang,a Yuanchang Li,a Zhongzhou Cheng,b Kai Xu,a Xueying Zhan,a

Zhenxing Wanga and Jun He*a

3D semiconductor nanostructures have proved to be a rich system for the exploring of high-performance

pseudocapacitors. Herein, a novel 3D WO3 nanotree on W foil is developed via a facile and green method.

Both capacitance and conductivity of the WO3 nanotree electrode are greatly improved after hydrogenation

treatment (denoted as H-WO3). First-principles calculation based on the experiments reveals the

mechanism of the hydrogenation treatment effect on the 3D WO3 nanotrees. The surface O of 3D WO3

nanotrees gains electrons from the adsorbed H, and consequently certain electrons are back-donated

to the neighboring W, thus providing the conducting channel on the surface. Ultrathin V2O5 films were

coated on the H-WO3 nanotrees via a simple, low-cost, environmentally friendly electrochemical

technique. This V2O5/H-WO3 electrode exhibited a remarkable specific capacitance of 1101 F g�1 and an

energy density of 98 W h kg�1. The solid-state device based on the V2O5/H-WO3 electrodes shows

excellent stability and practical application. Our work opens up the potential broad application of

hydrogenation treatment of semiconductor nanostructures in pseudocapacitors and other energy

storage devices.

Introduction

Pseudocapacitors with high-performance are of great scientificand practical importance due to rapidly growing global energyconsumption.1,2 As one kind of significant supercapacitor,pseudocapacitors are dominated by Faradaic reactions onelectrode materials. In this regard, transition metal oxidesand hydroxides are promising pseudocapacitive materials becauseof their good electrochemical reversibility, high rate capacitancein aqueous electrolytes and high specific capacitance.3–5 Recently,cobalt based materials (CoO,6 Co(OH)2,7 Co3O4

8,9 etc.), MnO210,11

and V2O512–14 have attracted growing interest.15–18 Vanadium

pentoxide (V2O5) is a promising metal oxide owing to its stablecrystal structure, high Faradaic activity, and wide potentialwindow with both aqueous and organic electrolytes.12–14

However, the limited ion diffusion within conventional denseelectrode film and poor electron transfer properties hinder itstheoretical performance.19,20 In order to overcome this defi-ciency, a composite with other conductive materials, such ascarbon nanomaterials,21,22 conducting polymers,6,23 metal nano-particles,10,24 and some semiconductors (ZnO,25 TiO2

11,15) hasbeen applied to enhance the conductivity of the pseudocapaci-tive oxide materials. Among these materials, semiconductornanostructures have recently been extensively studied as scaf-folds and exhibited promising properties. Meanwhile, 3D nano-structures with high electroactive surface area have proved tobe a rich system for the exploring of high-performance pseudo-capacitors. To date, a large number of high-surface-areacarbonaceous materials12 and freeze-drying processes13 basedon V2O5 nanostructure have been used to construct 3D archi-tectures to improve electrochemical performance advantageousfor supercapacitors. Despite these achievements, the fabrica-tion of cost-efficient, high performance semiconductor-basedelectrodes still remains a challenge. This may be due to thedifficulties in synthesis of low cost, high conductivity 3Dsemiconductor nanostructures.

As mentioned above, high electric conductivity plays acritical role in an efficient pseudocapacitor. Hydrogen treat-ment has been used to introduce oxygen vacancies into somepotential candidates like TiO2

11,15 and ZnO,25 and many research

a National Center for Nanoscience and Technology, Beijing 100190, P. R. China.

E-mail: [email protected] School of Materials Science and Engineering, University of Science and Technology

Beijing, Beijing 100083, P. R. China

† Electronic supplementary information (ESI) available: SEM image (Fig. S1) andhigh-resolution O1s XPS spectra (Fig. S2) of the WO3 nanotrees on W foil afterhydrogenation; the calculation equations of areal capacitance, specific capaci-tance, volumetric capacitance, power and energy density (eqn (S1)–(S6)); relatedelectrochemical measurements of electrodes (Fig. S3–S5 and Table S1). See DOI:10.1039/c4cp01200c

Received 20th March 2014,Accepted 17th April 2014

DOI: 10.1039/c4cp01200c

www.rsc.org/pccp

PCCP

PAPER

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article OnlineView Journal | View Issue

This journal is© the Owner Societies 2014 Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 | 12215

workers11,15,25 have reported the improved electrical conduc-tivity of above metal oxide by hydrogenation. Moreover, thehydroxyl groups introduced on the metal oxide surface duringhydrogenation can modify the electrochemical activity andincrease the pseudocapacitance of metal oxide electrodes.15,20,26

However, the mechanism governing the enhanced conductivityand hydroxyl groups during hydrogen treatment remainsunclear.

Herein, a novel 3D WO3 nanotree structure on W foil isdeveloped via a facile, environmentally friendly and low-costmethod. We design and fabricate a hybrid architecture bycoating ultrathin V2O5 film on hydrogenated WO3 (denoted asH-WO3) nanotrees. After hydrogen treatment, the carrierconcentration of WO3 nanotrees can be increased by 2 ordersof magnitudes, and thus, without the introduction of otherconductive materials (such as, noble metal or polymers) andcomplex processes, the conductivity of H-WO3 nanotrees isgreatly improved. Based on the experimental results, theoreticalsimulation and calculation reveal the mechanism of the hydro-gen treatment effect on the 3D WO3 nanotrees. In addition, theconductive W foils could directly serve as efficient currentcollectors reducing the ‘‘dead volume’’ of capacitors. The super-capacitors based on H-WO3 nanotrees and hybrids with ultra-thin V2O5 film exhibit high electrochemical performance. Thespecific capacitance (1101 F g�1) of the V2O5/H-WO3 electrode isabout 4 times larger than that (288 F g�1) of the V2O5/WO3

electrode. More significantly, all-solid-state supercapacitor (SC)devices fabricated with our electrodes demonstrate excellentperformance such as good specific capacitance, long lifetime,high energy density and operation voltage.

ExperimentalPreparation of H-WO3 nanotrees on W foil

WO3 nanotrees were fabricated on tungsten metal foil by asimple hydrothermal method. Detailedly, the W foils annealedin air at 500 1C for 30 min were placed in a 50 mL autoclavewith concentrated nitric acid (313 mL), H2C2O4�2H2O (1.56 g)and Rb2SO4 (0.2 g) and treated at 150 1C for 30 h. After thehydrothermal reaction, the foils washed with deionized waterand dried at room temperature were annealed in the furnace inair at 500 1C for 30 min. The H-WO3 nanotrees were obtained byannealing the WO3 nanotrees in a hydrogen atmosphere atdifferent temperatures (250, 350, 450, 550 1C).

Electrodeposition of ultrathin V2O5 film on H-WO3 nanotrees

The electrodeposition of V2O5 film was performed using athree-electrode system through the cyclic-voltammetry (CV)method employing a CHI 660D electrochemical workstation(Chenhua, Shanghai). In this system, a piece of W foil (1 �1.5 cm2) was used as a working electrode with an Ag/AgClelectrode as a reference electrode and a Pt wire as the counterelectrode. The electrolyte was prepared by dissolving 0.50 gVOSO4 in 60 mL deionized water with the pH of 1.8 by adding1 M H2SO4. Then, the deposition was performed in a potential

window of 0.555 V to 1.555 V. During the deposition, theelectrolyte was heated at 75 1C on a hot plate. After deposition,samples were annealed at 200 1C under ambient conditions for12 h. The mass of the V2O5 materials was measured using anelectronic balance (BT 125D) with �10 mg accuracy. The loadingmass of V2O5 film was 0.86 mg cm�2.

Assembly of the solid-state supercapacitor

The solid device was assembled by using two pieces of V2O5/H-WO3 electrodes with a separator (NKK TF40, 35 mm) andpolyvinyl alcohol (PVA)–LiCl gel as a solid electrolyte. A PVA–LiClelectrolyte was prepared by mixing LiCl (12.6 g) and PVA (6 g) in60 mL deionized water. The whole mixture was heated up to85 1C under vigorous stirring until the solution became clear for2 h. Then, two pieces of electrodes and the separator were soakedin the gel for about 5 min and then assembled together. Finally,the device was kept at 60 1C for 12 h to remove the excess water inthe electrolyte. After the electrolyte became hard, the solid-stateSSC device was prepared with a thickness of around 0.8 mm.

Material characterization and electrochemical measurements

The structure of the as-prepared samples was examined using aHitachi S-4800 field-emission scanning electron microscope(SEM) equipped with an X-ray energy dispersive spectrometer(EDS), a transmission electron microscope (TEM, JEM-2100F),an X-Ray diffractometer (XRD, Philips X’Pert Pro Super, Cu-Karadiation with l = 1.5418 Å) and an X-ray photoelectron spectro-scope (XPS, ESCALAB250Xi). Raman spectra of the sampleswere obtained using an InVo-RENISHAW system. The mass ofV2O5 film was measured on an electronic balance (BT 125D)with �10 mg accuracy. The electrochemical properties of thesamples were investigated with cyclic voltammetry (CV), electro-chemical impedance spectroscopy (EIS), Mott–Schottky and galvano-static charge/discharge (GCD) measurements in a conventionalthree-electrode cell employing a CHI 660D electrochemical work-station (Chenhua, Shanghai) and an EG&G Princeton AppliedResearch VMP3 workstation.

Results and discussion

The growth procedures of H-WO3 electrodes are illustrated inFig. 1a. Aligned WO3 nanotrees are directly grown on W foil by amodified hydrothermal process.27 The as-prepared WO3 nano-trees are annealed in a hydrogen atmosphere at differenttemperatures (250–550 1C) for 30 min for the creation of oxygenvacancies in WO3 nanotrees. As shown in SEM images in Fig. 1b(top view) and Fig. 1c (side view), WO3 nanotrees are verticallyaligned on the W foil with a height of B2 mm. Fig. 1d shows thetypical TEM image of a WO3 nanotree. The branches are grownalong the hexagonal-symmetry axis on the side face of thetrunks. Both trunks and branches were single crystals, withthe long-length axes oriented toward the h001i direction. Beforehydrogenation, high-resolution transmission electron micro-scopy (HRTEM) of the WO3 nanotrees illustrated in Fig. 1eshows parallel lattice fringes with the spacing of 0.380 nm

Paper PCCP

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article Online

12216 | Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 This journal is© the Owner Societies 2014

corresponding to the (002) plane. The corresponding selectedarea electron diffraction (SAED) (Fig. 1f ) pattern confirms thatthe WO3 nanotrees have crystalline quality and grow along the[001] axis. While, after the treatment in a hydrogen atmosphere,the color of WO3 nanotrees has changed, which suggestspossible modification in crystal structure or phase change. Asis shown in Fig. 1g, the edge of WO3 nanotrees is found to beamorphous and disordered which probably result from thehydrogenation of WO3 nanotrees. Meanwhile, the core of H-WO3

nanotrees still keeps high crystalline quality.XRD spectra shown in Fig. 2a reveals the crystalline struc-

tural difference between pristine and hydrogen treated WO3

nanotrees. After subtracting the diffraction peaks originatingfrom W foil ((200), JCPDS card No. 04-0806),28 the peaks collectedfrom pristine WO3 can be indexed as hexagonal WO3 (h-WO3,JCPDS Card No. 85-2460). The emergence of new peaks in thesamples hydrogen treated at a temperature of above 450 1C showsthe existence of the WO2.9 phase (JCPDS card No. 05-0386) andoxygen vacancies. Furthermore, Raman analysis (Fig. 2b) alsoconfirms the conversion of h-WO3 to W2.9 during hydrogentreatment.29 The characteristic Raman peaks of WO3 tend tobroaden as the annealing temperature increases, which isexpected for the increased amount of oxygen vacancies.

XPS is an effective measurement to examine the effect ofhydrogenation on the chemical composition and the oxidationstate of WO3 nanotrees. Fig. 2c shows the normalized W 4f corelevel XPS spectra of WO3 and H-WO3 at 450 1C and 550 1C. Twobroad peaks centred at B37.7 eV and B35.5 eV correspond tothe characteristic W 4f5/2 and W 4f7/2. The W 4f peaks of H-WO3

(at 450 1C and 550 1C) are slightly broader than that of pristine WO3

with a shoulder at the lower binding energy region (B37.0 eV and34.3 eV). This difference also confirms the presence of W5+ (oxygenvacancies) in the hydrogenated WO3 nanotrees with the increase ofannealing temperature. The result suggests that oxygen vacanciesare created in H-WO3 nanotrees during hydrogenation.29–32

The amount of W5+ in the H-WO3 sample increased with theannealing temperature from 450 1C to 550 1C. This resultsuggests the much more oxygen vacancies in H-WO3 treatedat 550 1C than that at 450 1C. In addition, the O 1s core levelXPS spectra of WO3 and H-WO3 nanotrees are compared inFig. S2 (ESI†). The presence of the broader shoulder at 531.7 eVof H-WO3 (550 1C) nanotrees is assigned to the formation ofhydroxyl (O–H), indicating the hydroxyl groups introduced onthe metal oxide surface during hydrogenation.16,29,31

The foregoing results confirm that the hydrogenation tem-perature is expected to be influential on the amount of oxygenvacancies and hydroxyl groups. As some results reported,11,15,25

oxygen vacancies and hydroxyl groups are important to electro-chemical performance of metal oxides. Fig. 3a displays the CVscollected for the H-WO3 electrodes at different hydrogenatedtemperatures with potential windows ranging from �0.4 V to0.4 V. When the hydrogenation temperature increases from250 1C to 550 1C, the capacitive current density of the H-WO3

electrode is obviously increased as compared to that of the WO3

electrode. The H-WO3 electrode hydrogenated at 450 1C shows thebest capacitive property possibly owing to apposite the amount ofoxygen vacancies and hydroxyl groups, which can be furtherconfirmed by our calculation. As expected, the H-WO3 samplehydrogenated at 450 1C yields the highest areal capacitancecalculated through eqn (S1) (ESI†) of 106.5 mF cm�2 at a scanrate of 50 mV s�1 (Fig. 3b). Enlightened by the reduction ofelectrical conductivity resulted from over much hydroxyl groupsand hydrogen embrittlement of W foil,15 we test EIS of differentelectrodes (Fig. 3c). From the point intersecting with the real axisin the range of high frequency,33,34 the internal resistance ofH-WO3 at 450 1C is 4.37 O, manifesting the lowest internal resistanceamong all electrodes. While the H-WO3 electrode hydrogenated at550 1C exhibits a higher internal resistance of 16.51 O, confirmingthat too high hydrogenated temperature is not favorable for electrical

Fig. 1 (a) A schematic diagram showing the fabrication of H-WO3 nano-trees on W foil. SEM image of WO3 nanotrees on W foil (b) top view and(c) side view. (d) Low magnification TEM image of a single WO3 nanotree.(e) HRTEM image of a typical WO3 nanotree. (f) SAED of the selected areain (e). (g) TEM image of H-WO3 nanotrees (treated at 450 1C) (the inset isHRTEM of the edge of the nanotree).

Fig. 2 (a) XRD spectra of pristine WO3 and H-WO3 samples at tempera-tures of 250, 350, 450, 550 1C. The diffraction peaks of WO2.9 are high-lighted by *. (b) Raman spectra of WO3 and H-WO3 prepared at varioustemperatures. (c) High-resolution W4f XPS spectra of pristine WO3 andWO3 prepared at 450 and 550 1C.

PCCP Paper

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article Online

This journal is© the Owner Societies 2014 Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 | 12217

conductivity of WO3 nanotrees. All CV curves of the H-WO3

electrode hydrogenated at 450 1C at various scan rates in Fig. 3ddemonstrate the ideal capacitive behaviours and high ratecapabilities of this electrode. The GCD curves at various currentdensities (Fig. S3a, ESI†) and the stable CV curves (Fig. S3b,ESI†) also indicate excellent pseudocapacitive performance ofthe H-WO3 electrode hydrogenated at 450 1C.15,35,36

Mott–Schottky (MS) measurement is a good method to inves-tigate the electrical properties based on the Schottky barrierformation between the semiconductor materials and an electro-lyte.11,15,29,37 As demonstrated in Fig. 4a, capacitances are derivedfrom EIS obtained at each potential with 1 KHz frequency in thedark. The WO3 and H-WO3 electrodes (inset of Fig. 4a) exhibitpositive slopes, indicating n-type semiconductor character.Carrier densities (Nd) of WO3 samples are calculated by thefollowing eqn (1). Where Nd is the donor density, e is therelative dielectric constant of WO3 (e = 20), e0 is the permittivityof a vacuum, V is the potential applied at the electrode, and A isthe area in contact with the electrolyte.

Nd ¼2

ee0eA2

dV

dCs�2 (1)

The carrier densities of untreated WO3 and H-WO3 nanotreesare calculated to be 2.04 � 1011 cm�3 and 1.64 � 1013 cm�3,respectively. Although the MS equation is derived from a planarmodel, these results are still qualitatively comparable as theseelectrodes have similar morphology and surface area. Typicalgalvanostatic charge/discharge curves of different WO3 electrodescollected at a current density of 0.17 mA cm�2 are shown in Fig. 4b.The areal capacitance of the H-WO3 sample is 104.74 mF cm�2,which is about 1.6 times that of the WO3 sample (65.67 mF cm�2).

Noteworthily, it shows a smaller IR drop (0.001 V), confirming theenhanced electrical conductivity of the H-WO3 electrode.

First-principles calculations are carried out to study theeffect of hydrogenation on WO3 nanotrees. Whether terminatedby W or O atoms (see Fig. 4c), the pristine WO3 nanowires alwayshave some free carriers because of the unsaturated defect states atthe surface from the first-principles calculations. However, thesestates are highly chemically active and easily passivated by theambient gas. Our results show that the H binding energies tounsaturated W and O are as high as 4.4 and 6.2 eV, respectively.This reveals that the surface conduction due to defect states shouldbe significantly suppressed and hence a bad metallic behaviourgiven the fabricated condition of our WO3 nanowires. To illustratethe hydrogen-induced conduction enhancement, we here use theW-terminated nanowire (black dashed circle in Fig. 4c) for discus-sions while the physics is similar for O-terminated ones. Note thatwe preliminarily passivate the surface W atoms by H to mimic theexperimental conditions. For hydrogenation, H prefers the bridge-site O that links two surface W atoms with the binding energyof 3.4 eV. Fig. 4d demonstrates the corresponding electronicstructure. Obviously, the system presents a large number ofn-type free carriers near the Fermi level, consistent with our MSmeasurements, responsible for the increased metallicity. Themechanism for enhanced conductivity is the same as that inother hydrogenated oxides,38,39 i.e., redistribution of W electronicstates at the surface. In the inset, we show the charge redistributionupon H adsorption. The electron depletion occurs around the H.The surface O gains electrons from the adsorbed H, and conse-quently certain electrons are back-donated to the neighboring W,

Fig. 3 (a) CV curves collected at a scan rate of 100 mV s�1 for H-WO3

nanotrees hydrogenated at various temperatures. (b) Areal capacitancescalculated for H-WO3 electrodes as a function of scan rate. (c) Nyquistplots for pristine WO3 and hydrogenated WO3 electrodes at temperaturesof 250, 350, 450, 550 1C. (d) CV curves of H-WO3 (prepared 450 1C) atdifferent scan rates from 10 to 200 mV s�1. The effective area of theelectrode is 1.5 cm2.

Fig. 4 (a) Mott–Schottky plots collected for WO3 and (450 1C) H-WO3

samples. (b) GCD curves of WO3 and H-WO3 (450 1C) electrodes at acurrent density of 0.17 mA cm�2. The effective area of the electrode is1.5 cm2. (c) Top view of a 4 � 4 � 1 WO3 supercell with hexagonalstructure. Black and blue dashed circles denote the nanowires terminatedby W and O atoms. The purple and red balls correspond to W and O atoms,respectively. (d) Total/local density of states of a hydrogenated WO3

nanowire with W termination. In the calculation, the surface W is passi-vated and the hydrogenation occurs at the bridge O sites. See the text fordetails. The inset shows the electron density difference before and afterhydrogenation projected to the adsorbed H plane. The blue areas indicateelectron loss while the red areas indicate electron gain.

Paper PCCP

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article Online

12218 | Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 This journal is© the Owner Societies 2014

thus providing the conducting channel on the surface.This agrees well with the dominant feature of W d-states nearthe Fermi level. The back-donated electrons also lead to thedecrease of the nominal valence of W (+5) as observed in ourXPS studies.

The aforementioned experimental and calculation resultsconvincingly confirm that H-WO3 (450 1C) nanotrees on W foilhave better electrochemical performance and electrical proper-ties. Here, the V2O5/H-WO3 hybrid electrode is fabricated via anelectrodeposition process. SEM images in left of Fig. 5a showthat the H-WO3 nanotrees are completely filled with ultrathinV2O5 film after 15 cycles of CV deposition (Fig. S4a, ESI†). Thedetailed microstructures of the V2O5/H-WO3 hybrid are furthercharacterized by TEM. As is demonstrated in right of Fig. 5a,the V2O5 film covered uniformly on the H-WO3 nanotrees. Noclear fringe spacing is observed for the V2O5 outer layer (inset ofFig. S4a, ESI†), indicating that the V2O5 film is amorphous.Additionally, the local energy dispersive X-ray spectroscopy(EDS) analysis also confirms the existence of W, O, V elementsin our hybrid structure (Fig. 5b). The Raman scattering spectrashown in Fig. 5c clearly display the polycrystalline nature of theV2O5 film and broader WO3 peaks because of hydrogen treat-ment. From the XPS curves, where the characteristic satellitesof V5+ 2p3/2 and 2p1/2 bands located at the binding energies of517.5 eV and 525 eV are presented, it is obviously known thatultrathin V2O5 film formed in our architecture.13

To demonstrate the electrochemical performance of compo-site structures, the electrochemical properties of samples arefirst characterized in a three-electrode system with CVs, GCD,and EIS measurements in 1 M Na2SO4 aqueous solution. TheCVs and EIS of bare WO3, H-WO3 and V2O5/WO3 electrodes areshown in Fig. S4b and c (ESI†), revealing that the V2O5/H-WO3

electrode shows best electrochemical performance and lowestequivalent series resistance (ESR) among electrodes. The GCD

measurements shown in Fig. 6a confirm this result. The arealcapacitance (Ca) and specific capacitance of the electrodes arecalculated by eqn (S2) and (S3) (ESI†).4,15 The areal capacitanceof the V2O5/H-WO3 electrode (Fig. 6b and Fig. S4d, ESI†) isdetermined to be 954 mF cm�2 (1101 F g�1, only the mass ofV2O5) at a current density of 0.17 mA cm�2, which is nearlythree times that of the V2O5/WO3 sample (250 mF cm�2) andhigher than the value of reported 3D V2O5 nanosheets.13

Fig. 5 (a) SEM images the H-WO3 nanotrees after electrodeposition of V2O5

film (upper left), the high magnification SEM image of V2O5 film on H-WO3

nanotrees (bottom left) and the TEM image of the V2O5/H-WO3 hybrid.(b) EDS pattern of V2O5/H-WO3 nanotrees. (c) Raman spectra of bulk WO3,bulk V2O5 and the hybrid V2O5/H-WO3 material. (d) XPS spectra of theV2O5/H-WO3 architecture, indicating the presence of V5+ in the sample.

Fig. 6 (a) GCD curves of WO3, H-WO3, V2O5/WO3 and V2O5/H-WO3

electrodes at a current density of 0.33 mA cm�2. (b) Areal capacitanceand specific capacitance (by account the mass of V2O5) of V2O5/WO3 andV2O5/H-WO3 electrodes as a function of the current density. (c) GCDcurves of the V2O5/H-WO3 electrode at various current densities.

PCCP Paper

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article Online

This journal is© the Owner Societies 2014 Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 | 12219

Meanwhile, the GCD curves of the V2O5/H-WO3 electrode arerecorded with various current densities and shown in Fig. 6c,demonstrating good capacitive behaviour.

Considering the feasibility for capacitance applications, anall-solid-state symmetric SC is assembled with two V2O5/H-WO3

electrodes using a solid PVA–LiCl gel electrolyte and a separatorbetween them shown in Fig. 7a and Fig. S5a (ESI†). The superiorperformance of the solid-state SSC device is confirmed by galvano-static charge/discharge measurements. As shown in Fig. 7b, thecharging and discharging curves of this solid device at differentcurrent densities are reasonably symmetric. A small slopingpotential profile with a potential plateau is observed in GCD curvesarising from the Faradaic reaction.10,40 The volumetric capacitancesat various current densities are calculated based on the volume ofall solid devices (eqn (S4), ESI†). As is demonstrated in Fig. S5b(ESI†), our device shows very good rate capacitance, with 80% ofvolumetric capacitance retained when the current density increasedfrom 0.17 (0.98 F cm�3) to 2 mA cm�2 (0.78 F cm�3). The long-termcycling performance of this device is evaluated through CV at ascan rate of 100 mV s�1 for 7000 cycles. As is shown in Fig. 7c, thecapacitance value remains nearly constant. The capacitance valuegradually increases at the first 1000 cycles, inferring that theelectrochemical property of our materials has an activation process,similar to some reported results.16

Furthermore, power density (P) and energy density (E) are twoindispensable parameters for evaluating the electrochemicalperformance of the supercapacitors. Fig. 7d displays the Ragoneplots of our device (eqn (S5) and (S6), ESI†). Based on the mass ofV2O5, the comparison of specific capacitance, energy density andpower density of different electrodes is presented in Table S1(ESI†).12–14,40,41 The thin film with 3D nanostructure of V2O5

improves the Na+ (Li+) loading due to the lower diffusion distanceand high surface area. Such superior capacitive performance of

our SSC solid-state device can be attributed to the followingmerits: (1) the pseudocapacitive hybrid nanotree materials havea direct contact to the current collector (W foil); (2) the hybridnanotrees are well immobilized to the substrate, reducing the‘‘dead volume’’ in electrode materials; (3) both the scaffoldH-WO3 and V2O5 thin film are good pseudocapacitive metaloxides. To demonstrate the potential application, two soliddevices were employed to power a red light-emitting-diode(LED) well for about 2 min after charging at 2 mA cm�2 for30 s (the inset of Fig. 7d).

Conclusions

In conclusion, a practical and cost-effective strategy has beendeveloped to fabricate the hybrid nanostructure arrays for pseudo-capacitor applications. 3D ordered H-WO3 nanotrees with excellentconductivity function as efficient scaffolds to support V2O5 film.Noteworthily, this V2O5/H-WO3 electrode shows outstandingelectrochemical performance in supercapacitors such as highspecific capacitance, good stability and high energy density.Combined with the advantages of low cost, easy operation andenvironmentally friendly nature, this novel solid-state deviceshows great promise for application in commercial pseudo-capacitors. Our work opens up ways to construct high-performancepseudocapacitive materials without using any carbon, noble metaland polymer-based conducting media. Moreover, these findingsexhibit the potential broad application of hydrogen treatment inpseudocapacitors and other energy storage devices.

Acknowledgements

This work was supported by the 973 Program of the Ministryof Science and Technology of China (No. 2012CB934103), the100-Talents Program of the Chinese Academy of Sciences(No. Y1172911ZX) and National Natural Science Foundationof China (21373065, 21307020).

Notes and references

1 A. S. Arico, P. Bruce, B. Scrosati, J. M. Tarascon andW. Van Schalkwijk, Nat. Mater., 2005, 4, 366–377.

2 P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.3 H. Chen, L. Hu, M. Chen, Y. Yan and L. Wu, Adv. Funct.

Mater., 2014, 24, 934–942.4 H. B. Li, M. H. Yu, F. X. Wang, P. Liu, Y. Liang, J. Xiao,

C. X. Wang, Y. X. Tong and G. W. Yang, Nat. Commun., 2013,4, 1894.

5 M. Vangari, T. Pryor and L. Jiang, J. Energy Eng., 2013, 139,72–79.

6 C. Zhou, Y. Zhang, Y. Li and J. Liu, Nano Lett., 2013, 13,2078–2085.

7 L. Cao, F. Xu, Y. Y. Liang and H. L. Li, Adv. Mater., 2004, 16,1853–1857.

8 L. Yang, S. Cheng, Y. Ding, X. Zhu, Z. L. Wang and M. Liu,Nano Lett., 2012, 12, 321–325.

Fig. 7 (a) Schematic diagram of a solid-state device with a separator andPVA–LiCl gel electrolyte. (b) GCD curves of the solid-state device atvarious current densities. (c) Cycling performance and volumetric capa-citance (based on volume of all solid device) of the device in 0–0.8 Vpotential windows at a scan rate of 100 mV s�1 for 7000 cycles. (d) Ragoneplot of this device measured with the PVA–LiCl gel electrolyte (the insetsare pictures of our device).

Paper PCCP

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article Online

12220 | Phys. Chem. Chem. Phys., 2014, 16, 12214--12220 This journal is© the Owner Societies 2014

9 J. P. Liu, Adv. Mater., 2011, 23, 2076–2081.10 X. Lu, T. Zhai, X. Zhang, Y. Shen, L. Yuan, B. Hu, L. Gong,

J. Chen, Y. Gao, J. Zhou, Y. Tong and Z. L. Wang, Adv.Mater., 2012, 24, 938–944.

11 X. Lu, M. Yu, G. Wang, T. Zhai, S. Xie, Y. Ling, Y. Tong andY. Li, Adv. Mater., 2013, 25, 267–272.

12 A. Ghosh, E. J. Ra, M. Jin, H.-K. Jeong, T. H. Kim, C. Biswasand Y. H. Lee, Adv. Funct. Mater., 2011, 21, 2541–2547.

13 J. Zhu, L. Cao, Y. Wu, Y. Gong, Z. Liu, H. E. Hoster, Y. Zhang,S. Zhang, S. Yang, Q. Yan, P. M. Ajayan and R. Vajtai, NanoLett., 2013, 13, 5408–5413.

14 S. D. Perera, B. Patel, J. Bonso, M. Grunewald, J. P. Ferrarisand K. J. Balkus, Jr., ACS Appl. Mater. Interfaces, 2011, 3,4512–4517.

15 X. H. Lu, Nano Lett., 2012, 12, 1690–1696.16 J. Kang, A. Hirata, H. J. Qiu, L. Chen, X. Ge, T. Fujita and

M. Chen, Adv. Mater., 2014, 26, 269–272.17 X. Xiao, Adv. Energy Mater., 2012, 2, 1328–1332.18 M. Toupin, T. Brousse and D. Belanger, Chem. Mater., 2004,

16, 3184–3190.19 Z. F. Li and E. Ruckenstein, Langmuir, 2002, 18, 6956–6961.20 D. Choi, G. E. Blomgren and P. N. Kumta, Adv. Mater., 2006,

18, 1178–1182.21 Z. Fan, J. Yan, T. Wei, L. Zhi, G. Ning, T. Li and F. Wei,

Adv. Funct. Mater., 2011, 21, 2366–2375.22 J. Yan, Z. Fan, W. Sun, G. Ning, T. Wei, Q. Zhang, R. Zhang,

L. Zhi and F. Wei, Adv. Funct. Mater., 2012, 22, 2632–2641.23 Q. Qu, Y. Zhu, X. Gao and Y. Wu, Adv. Energy Mater., 2012, 2,

950–955.24 Q. Lu, M. W. Lattanzi, Y. Chen, X. Kou, W. Li, X. Fan,

K. M. Unruh, J. G. Chen and J. Q. Xiao, Angew. Chem.,Int. Ed., 2011, 50, 6847–6850.

25 P. Yang, X. Xiao, Y. Li, Y. Ding, P. Qiang, X. Tan, W. Mai,Z. Lin, W. Wu, T. Li, H. Jin, P. Liu, J. Zhou, C. P. Wong andZ. L. Wang, ACS Nano, 2013, 7, 2617–2626.

26 X. Fan, Y. Lu, H. Xu, X. Kong and J. Wang, J. Mater. Chem.,2011, 21, 18753–18760.

27 M. Shibuya and M. Miyauchi, Adv. Mater., 2009, 21,1373–1376.

28 J. Zhang, X. L. Wang, X. H. Xia, C. D. Gu and J. P. Tu, Sol.Energy Mater. Sol. Cells, 2011, 95, 2107–2112.

29 G. Wang, Y. Ling, H. Wang, X. Yang, C. Wang, J. Z. Zhangand Y. Li, Energy Environ. Sci., 2012, 5, 6180–6187.

30 L. Cheng, Y. Hou, B. Zhang, S. Yang, J. W. Guo, L. Wu andH. G. Yang, Chem. Commun., 2013, 49, 5945–5947.

31 S. Bathe and P. Patil, Solid State Ionics, 2008, 179, 314–323.32 J. S. Lee, I. H. Jang and N.-G. Park, J. Phys. Chem. C, 2012,

116, 13480–13487.33 X. Wang, B. Liu, Q. Wang, W. Song, X. Hou, D. Chen,

Y. B. Cheng and G. Shen, Adv. Mater., 2013, 25, 1479–1486.34 L. Q. Mai, A. Minhas-Khan, X. Tian, K. M. Hercule,

Y. L. Zhao, X. Lin and X. Xu, Nat. Commun., 2013, 4, 2923.35 M. Salari, K. Konstantinov and H. K. Liu, J. Mater. Chem.,

2011, 21, 5128–5133.36 L. Gao, X. Wang, Z. Xie, W. Song, L. Wang, X. Wu, F. Qu,

D. Chen and G. Shen, J. Mater. Chem. A, 2013, 1, 7167–7173.37 I. n. Mora-Sero, F. Fabregat-Santiago, B. Denier, J. Bisquert,

R. n. Tena-Zaera, J. Elias and C. Levy-Clement, Appl. Phys.Lett., 2006, 89, 203117.

38 Y. Wang, B. Meyer, X. Yin, M. Kunat, D. Langenberg,F. Traeger, A. Birkner and C. Woll, Phys. Rev. Lett., 2005,95, 266104.

39 F. Lin, S. Y. Wang, F. W. Zheng, G. Zhou, J. Wu, B. L. Gu andW. H. Duan, Phys. Rev. B: Condens. Matter Mater. Phys., 2009,79, 266104.

40 Q. T. Qu, Y. Shi, L. L. Li, W. L. Guo, Y. P. Wu, H. P. Zhang,S. Y. Guan and R. Holze, Electrochem. Commun., 2009, 11,1325–1328.

41 H. Zhang, A. Xie, C. Wang, H. Wang, Y. Shen and X. Tian,ChemPhysChem, 2014, 15, 366–373.

PCCP Paper

Publ

ishe

d on

17

Apr

il 20

14. D

ownl

oade

d by

Nat

iona

l Cen

ter

for

Nan

oSci

ence

and

Tec

hnol

ogy,

Chi

na o

n 11

/08/

2015

08:

44:4

0.

View Article Online