Es

Xa

b

a

ARRAA

KARCSS

1

slcadtreiirolptgCoMc

0d

Synthetic Metals 159 (2009) 7–12

Contents lists available at ScienceDirect

Synthetic Metals

journa l homepage: www.e lsev ier .com/ locate /synmet

lectrochemical performance of RuOx/activated carbon black composite forupercapacitors

.J. Hea,∗, Y.J. Genga, S. Okeb, K. Higashib, M. Yamamotob, H. Takikawab

Department of Chemical Engineering, School of Chemistry and Chemical Engineering, Anhui University of Technology, 59 Hudong Road, Maanshan 243002, ChinaDepartment of Electrical and Electronic Engineering, Toyohashi University of Technology, Aichi 441-8580, Japan

r t i c l e i n f o

rticle history:eceived 17 May 2008eceived in revised form 1 July 2008ccepted 11 July 2008

a b s t r a c t

Hydrous ruthenium oxide/activated carbon black (RuOx/ACB) composite was synthesized for the firsttime for supercapacitors by a chemical impregnation method. The RuOx/ACB composite is characterizedby thermogravimetric analysis (TGA), field emission scanning electron microscope (FE-SEM) and trans-mission electron microscope (TEM). The results of TGA and TEM characterization showed that the reaction

vailable online 11 September 2008

eywords:ctivated carbon blackuthenium oxideomposite

temperature of composite moved to lower temperature due to the catalytic effect of RuOx. XRD characteri-zation of RuOx/AC illustrated that amorphous hydrous RuOx structure maintained in the composite heatedat 150 ◦C. The equivalent series resistance (ESR) of the composite almost increased with the increase ofRuOx content in the composite. The specific capacitance (Csp) of the composite increased with increasingRuOx content in the composite while the Csp of RuOx decreased from 1255.8 to 533.7 F g−1. The above

ilizat

tio[itp1

(teTts

2

2

pecific capacitanceupercapacitor

results showed that the ut

. Introduction

Supercapacitors have attracted increasing interest in the powerources due to its short charging time, high power density andong cycle life [1]. Supercapacitors included electric double layerapacitor (EDLC) and Faradic capacitor according to the charge stor-ge mechanism. The charge storage in an EDLC is based on theouble layer process at the electrode–electrolyte interface whilehe charge storage in a Faradic capacitor is based on the Faradicedox reactions. Ruthenium oxide (RuOx) is known as the bestlectrode material of the Faradic capacitor due to its outstand-ng specific capacitance (Csp) and long cycle life [2,3]. However,ts practical utilization was limited because of the high cost ofuthenium element. Thus, interest in research has focused on devel-ping a hybrid system that consists of an electrochemical doubleayer process and a Faradic redox reaction with RuOx/carbon com-osite materials. Carbon materials used as electrode materials inhe hybrid system included activated carbon [4–8], carbon aero-els [9], carbon nanotubes [10–13] and carbon black [14–16]. The

sp of supercapacitor relied on the content and microstructuref RuOx as well as the specific surface area of the composite.iller et al. [9] reported that the high annealing temperature of

a. 300 ◦C initiated a crystalline transition of RuO2 particle on

∗ Corresponding author. Tel.: +86 555 2311807; fax: +86 555 2311822.E-mail address: xjhe@ahut.edu.cn (X.J. He).

wstrfa

379-6779/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.synthmet.2008.07.008

ion of RuOx decreased with higher loading RuOx on ACB.© 2008 Elsevier B.V. All rights reserved.

he carbon aerogel and resulted in that the Csp of the compos-tes obtained by chemical vapor impregnation is lower than thatf the composites obtained by the sol–gel method. Zheng et al.17] found that RuO2 performance is highly sensitive to the anneal-ng temperature and that its structure changed from amorphouso crystalline above 175 ◦C. Zhang et al. [4] reported that RuO2/ACowers had the highest Csp when the powers were annealed at50 ◦C.

In this paper, hydrous ruthenium oxide/activated carbon blackRuOx/ACB) composite was synthesized by a chemical impregna-ion method and the effect of RuOx content in the composite on thequivalent series resistance (ESR) and the Csp were investigated.he aim of this research is to find the optimal RuOx content inhe composites for the maximum Csp of the electrode material ofupercapacitors.

. Experimental

.1. Preparation and characterization of the composite electrodes

Ruthenium precursors were impregnated onto ACB powdersith a specific surface area of 1270 m2 g−1 and primary particle

ize of 34 nm. The RuCl3 precursor solution was injected into a cuphat was previously charged with dried ACB powders under stir-ing by a magnetic bar. The ruthenium complex colloidal solutionormed with a slow addition of 1 M ammonium hydrogen carbon-te (NH4HCO3) solution into the aqueous RuCl3/ACB solution. The

8 ic Metals 159 (2009) 7–12

rbucwmRomp3haiasemtottp

X.J. He et al. / Synthet

esultant solution was filtered with the polyethersulfone mem-rane (0.2 �m) to get the ACB particles after being dispersed byltrasonic wave for 30 min. The composites with different RuOx

ontent were annealed in air at 150 ◦C for 2 h after being washedith distilled water and dried at room temperature. Thermogravi-etric analysis (TGA) for the composite with different loading

uOx were carried out in 30 ml min−1 air-flow at a heating ratef 5 ◦C min−1 in a thermal analysis instrument (TG-60, Japan). Theass of the composite in all cases was ca. 5 mg. The resultant com-

osite was mixed with binder (PTFE) in the weight ratio of 90:10 for0 min in an automatic mortar. The electrode (20 mm o.d., 1.2 mmigh) was dried at 120 ◦C for 2 h in a vacuum oven after cold presst 15 MPa pressure. The dried electrode (200 ± 20 mg) was soakedn 1 M H2SO4 solution in a vacuum oven for 30 min. The bare ACBnd heat treatment composites were characterized with field emis-ion scanning electron microscope (FE-SEM, Hitachi, S-4500) withnergy dispersive X-ray (EDX, IXRF Systems, EDS 2004) and trans-ission electron microscope (TEM, Tecnai G2 S-TWIN) with EDX. At

he same, the effect of annealed temperature on the microstructuref RuO2/activated carbon composite was introduced by charac-

erizing RuO2/activated carbon composites heated at differentemperature with X-ray diffraction (XRD, RINT-2500) for referenceurpose.

Fig. 1. Thermogravimetric analysis curves of ACB and RuOx/ACB composites.

2

actut1co

3

3

caRiaet

Fig. 2. FE-SEM images of the composite: (a) ACB and (b) RuOx2/ACB.

.2. Electrochemical measurements

The electrochemical experiments were carried out on automatic polarization system (Hokuto Denko, HZ-5000, Japan)onnected to a two-electrode test cell in 1 M H2SO4 solution withhe same two electrodes. Stainless foil and cellulose (TF4030) issed as the collector and separator, respectively. The electrode inhe cell was soaked in 1 M H2SO4 solution at −20 kPa pressure for0 min before each test. Cyclic voltagrammetry and discharge atonstant current were used to investigate the capacitive behaviorf the electrodes.

. Results and discussion

.1. Characterization of RuO2/ACB composites

Fig. 1 presents TGA curves of ACB and RuOx/ACB composites. Itan be seen from Fig. 1 that ACB is almost combusted up to 650 ◦C,nd the residual weight of RuOx/ACB composites is that of RuOx.uOx content in RuOx1/ACB, RuOx2/ACB and RuOx3/ACB composite

s 2.3, 4.5 and 9.6 wt.%, respectively. Moreover, the reaction temper-ture of composite moved to lower temperature due to the catalyticffect of RuOx, which could be observed at the beginning stage ofhe curves before 450 ◦C in Fig. 1.

ic Metals 159 (2009) 7–12 9

coifaEoptacNehahb0rR

X.J. He et al. / Synthet

Fig. 2 shows the FE-SEM images of the bare ACB and RuOx2/ACBomposites. It can be seen from Fig. 2a that the mean diameterf the pristine ACB maintains ca. 50 nm. Fig. 3 presents the TEMmages of the bare ACB and RuOx2/ACB composites. It can be seenrom Fig. 3a that the ACB is particle shape. A few small particlesppeared on the surface of the RuOx2/ACB composite in Fig. 3b. TheDS results in Fig. 3c proved that the small particle is composed ofxygen (O) and ruthenium (Ru) element. But, the particle shape isartly changed due to the catalysis effect of the loading RuOx onhe surface of ACB after the RuOx2/ACB composite was heated inir at 150 ◦C. Fig. 4 shows XRD patterns of RuOx/activated carbonomposite heated at different temperature for reference purpose.o peaks appeared in hydrous RuOx/AC heated at 150 ◦C while sev-ral diffraction peaks of RuO2 were observed in RuOx/AC compositeeated at 220 and 280 ◦C, as shown in Fig. 4. It can be deduced thatmorphous hydrous RuOx structure maintained in the composite

◦

eated at 150 C [4]. The above results agreed with those reportedy Kim [14]. Ru/O atomic ratio in the composites increases from.02 to 0.12 with increasing the RuO2·xH2O/ACB mass ratio in theaw materials. Thus, the small particles in Fig. 3b should be hydrousuOx crystallites.

Fig. 3. TEM images of the composite: (a) ACB; (b) RuOx2

Fig. 4. XRD patterns of RuOx/AC composite heated at different temperature.

/ACB; (c) EDX spectrum of RuOx2/ACB composite.

1 ic Metals 159 (2009) 7–12

3m

3e

ecpriar

Foatteb

E

wcd

Ftc

Ft

3

0 X.J. He et al. / Synthet

.2. Cyclic voltammetry and discharge behavior of the electrodeaterials

.2.1. Effect of discharge current on the ESR of the bare ACBlectrode

Fig. 5a presents cyclic voltammograms (CVs) of the bare ACBlectrode at different scan rate. Rectangular curves in Fig. 5a indi-ate the domination of double layer formation in the energy storagerocess. The curve shape become quasi rectangular at higher scanate, which is due to the resistance for ion migration in microporesn the composite electrode because that increasing the scan rateggravates the delay of the current to reach a horizontal value aftereversal of the potential scan.

Chronopotentiograms of the bare ACB electrode are shown inig. 5b. It can be seen from Fig. 5b that there is a sudden dropf potential at the beginning stage of discharge, which is associ-ted with the ohmic resistance of the cell. This result indicatedhe significant resistance on the ACB electrode. At the same time,he electrode potential is linearly dependent on the discharge time,xhibiting the capacitive behaviors. The ESR of the capacitor coulde evaluated by the following equation [1]:

SR = Vcharge − Vdischarge

2I(1)

here Vcharge is the cell voltage at the end of charge, Vdischarge is theell voltage at the beginning of the discharge, and I is the regulatedischarge current.

ig. 5. (a) CVs of the bare ACB electrode at different scan rate in 1 M H2SO4 solu-ion and (b) chronopotentiograms of the bare ACB electrode at different dischargeurrent.

e

EcitcRp

3e

w

C

wv

Fc

ig. 6. Effect of RuOx content in the composite on the ESR of the composite elec-rodes.

.2.2. Effect of RuOx content on the ESR of the compositelectrodes

Fig. 6 shows the effect of RuOx content in the composites on theSR of the electrodes at 25 mA discharge current in 1 M H2SO4. Itan be seen from Fig. 6 that the ESR of the composite electrodesncreased from 0.84 to 1.56 with the increase of RuOx content inhe composite of the electrodes. The increase of the ESR of theomposite electrodes was ascribed to the increase of amorphousuO2·xH2O content with the increase of RuOx content in the com-osite electrode because amorphous hydrous RuO2 is a dielectric.

.2.3. Effect of RuOx content on the Csp of the compositelectrodes

The Csp of the composite electrodes is shown in Fig. 7. The Csp

as calculated according to the following equation [18]:

sp = 2i

(�V /�t)m(2)

here Csp is the specific capacitance of the electrode, i is the specificoltammetric charge within the working voltage, �V is the dis-

ig. 7. Effect of RuOx content in the composite on the specific capacitance of theomposite electrodes.

ic Metals 159 (2009) 7–12 11

cottstRiiirbbtreacrnhntcca

C

wratwC

orf

R

arwC9ttscrtreiitdotioe

ud

4

aopccReictidtA

A

FstUCJ

R

X.J. He et al. / Synthet

harge voltage difference between the �t period, and m is the massf the active materials in one electrode. The factor of 2 comes fromhe total capacitance measured from the test cells being the addi-ion of two equivalent single-electrode capacitors in series. It can beeen from Fig. 7 that the Csp of the ACB electrode is 111.8 F g−1 whilehe Csp of hydrous RuOx1/ACB composite electrode with 2.3 wt.%uOx is 138.1 F g−1, an increase of ca. 23.5%. The Csp of the compos-

te electrode slowly increased when RuOx content in the compositencreased to 9.6 wt.%. As we all know, the total Csp of the compos-te electrodes was the sum of both double layer capacitance andedox capacitance. The capacitance contributed by the electric dou-le layer decreased because more surface area of ACB was coveredy RuO2 particles [19,20]. The competing factors caused that theotal Csp increased fast when RuOx content in the composite waselatively low. Further increase of RuOx loadings in the compositelectrode resulted in a slow increase when RuOx content was rel-tively high in the composite [4]. It was shown that the specificapacitance of the composite increased with the RuOx content, andeached 152.3 F g−1 when the RuOx content was 9.6 wt.%. Ruthe-ium is a noble metal and the cost of ruthenium precursor is quiteigh and the utilization rate of RuOx is an important parameter. Aew parameter (CRuOx

sp ) has been introduced to evaluate the utiliza-ion rate of RuOx [6]. This parameter is calculated by dividing theapacitance delivered from RuOx content by its weight. The specificapacitance of the mixture of ACB and RuOx has been calculatedccording to the following equation [21]:

mixture = m1CACBsp + m2CRuOx

pp

m1 + m2(3)

here CACBsp and CRuOx

sp are the specific capacitances of ACB and RuOx,espectively, and m1 and m2 are the mass in percentage of ACBnd RuOx in the composite. It was found that the capacitance con-ributed from RuOx decreased when the RuOx content increased,hich was consist with the conclusion of Su [22]. The calculated

RuOxsp was ca. 1255.8 F g−1 when the CRuOx

sp content was 2.3 wt.%.The electrochemical reaction of RuO2·xH2O is similar to that

f RuO2, which means both of them can be oxidized and reducedeversibly through the electrochemical protonation, shown in theollowing equation [23]:

uO2 + ıH+ + ıe− ↔ RuO2−ı(OH)ı, 0 ≤ ı ≤ 2 (4)

So, the change in capacitance characterization of RuOx isttributed to the concept that an amorphous structure of RuOx

etains the facile pathways for electron and proton conductionhile crystalline RuOx decreases the proton transport rate. The

RuOxsp decreased to 533.7 F g−1 when the CRuOx

sp content increased to.6 wt.%. This result probably originated from an incomplete reac-ion of RuOx. It was possible that the redox reaction started fromhe surface of RuOx nanoparticles and the reaction rate becamelower as the reaction proceeded. Thus, only smaller RuOx particlesould be fully reacted while larger RuOx particles remained partiallyeacted. The same phenomenon has been found in other litera-ure [24]. So the heavily loaded composite exhibited a significantlyeduced utilization of ruthenium oxide. The relationship of thelectrolyte impedance with the frequency can also give importantnformation on the electrochemical system [25]. The electrolytempedance curves of the supercapacitor tested in the same sys-em are shown in Fig. 8. It is shown that the electrolyte impedanceecreases with increasing frequency and an impedance plateau is

bserved in the high frequency range. The electrolyte impedance ishe lowest for ACB electrode at the high frequencies. The electrolytempedance of the RuOx3/ACB composite is always bigger than thatf the RuOx1/ACB composite. The above results illustrate that thelectrolyte impedance of the composite with higher loading RuOx is

[[

[

Fig. 8. Electrolyte impedance versus frequency for the composite electrodes.

sually bigger than that of the composite with lower loading RuOx

ue to the resistance of dielectric RuOx.

. Conclusions

Hydrous RuOx/ACB composite were successfully prepared bychemical impregnation method for supercapacitors. The results

f TGA and TEM characterization showed that the reaction tem-erature of composite moved to lower temperature due to theatalytic effect of RuOx and the shape of ACB particle was partlyhanged. XRD characterization illustrated that amorphous hydrousuOx structure maintained in the composite heated at 150 ◦C. Thelectrochemical results presented that the ESR of the compos-te almost increased with the increase of RuOx content in theomposite due to the increase of dielectric hydrous RuOx con-ent in the composite. The Csp of the composite increased withncreasing RuOx content in the composite while the Csp of RuOx

ecreased from 1255.8 to 533.7 F g−1. The above results showedhat the utilization of RuOx decreased with higher loading RuOx onCB.

cknowledgements

This work was partly supported by Natural Science Researchoundation and the Foundation for Youth Teacher in Univer-ity of Anhui Province in China (KJ2008A120, 2008JQ1026ZD),he Fund of Provincial Innovative Group for Processing & Cleantilization of Coal Resource and Research Project of Researchenter for Future Vehicle, Toyohashi University of Technology in

apan.

eferences

[1] Y.G. Wang, Y.Y. Xia, J. Electrochem. Soc. 153 (2006) A450.[2] X.R. Liu, P.G. Pickup, J. Power Sources 176 (2008) 410.[3] W.C. Fang, M.S. Leu, K.H. Chen, L.C. Chen, J. Electrochem. Soc. 155 (2008) K15.[4] J.R. Zhang, D.C. Jiang, B. Chen, J.J. Zhu, L.P. Jiang, H.Q. Fang, J. Electrochem. Soc.

148 (2001) A1362.[5] M. Ramani, B.S. Haran, R.E. White, B.N. Popov, J. Electrochem. Soc. 148 (2001)

A374.[6] J.H. Jang, S. Han, T. Hyeon, S.M. Oh, J. Power Sources 123 (2003) 79.[7] W.C. Chen, C.C. Hu, C.C. Wang, C.K. Min, J. Power Sources 125 (2004) 292.

[8] C.C. Hu, W.C. Chen, Electrochim. Acta 49 (2004) 3469.[9] J.M. Miller, B. Dunn, T.D. Tran, R.W. Pekala, J. Electrochem. Soc. 144 (1997) L309.10] X. Qin, S. Durbach, G.T. Wu, Carbon 42 (2004) 451.11] G.H. Deng, X. Xiao, J.H. Chen, X.B. Zeng, D.L. He, Y.F. Kuang, Carbon 43 (2005)

1566.12] J.K. Lee, H.M. Pathan, K.D. Jung, O.S. Joo, J. Power Sources 159 (2006) 1527.

1 ic Met

[[[[[[[

[[

2 X.J. He et al. / Synthet

13] I.H. Kim, J.H. Kim, Y.H. Lee, K.B. Kim, J. Electrochem. Soc. 152 (2005) A2170.14] H. Kim, B.N. Popov, J. Power Sources 104 (2002) 52.

15] G.Y. Yu, W.X. Chen, Y.F. Zheng, J. Zhao, X. Li, Z.D. Xu, Mater. Lett. 60 (2006) 2453.16] M. Min, K. Machida, J.H. Jang, K. Naoi, J. Electrochem. Soc. 153 (2006) A334.17] J.P. Zheng, P.J. Cygan, T.R. Jow, J. Electrochem. Soc. 142 (1995) 2699.18] Y.R. Nian, H. Teng, J. Electrochem. Soc. 149 (2002) A1008.19] Y. Sato, K. Yomogida, T. Nanaumi, K. Kobayakawa, Y. Ohsawa, M. Kawai, Elec-

trochem. Solid State Lett. 3 (2000) 113.

[[[

[

als 159 (2009) 7–12

20] C.C. Wang, C.C. Hu, Mater. Chem. Phys. 83 (2004) 289.21] H.F. Li, R.D. Wang, R. Cao, Micro. Mesop. Mater. 111 (2008) 32.

22] Y.Y. Su, F. Wu, L.Y. Bao, Z.H. Yang, N. Carbon Mater. 22 (2007) 53.23] Y.Y. Liang, H.L. Li, X.G. Zhang, J. Power Sources 173 (2007) 599.24] X.P. Dong, W.H. Shen, J.L. Gu, L.M. Xiong, Y.F. Zhu, H. Li, J.L. Shi, J. Phys. Chem. B

110 (2006) 6015.25] X.M. Liu, R. Zhang, L. Zhang, D.H. Long, W.M. Qiao, J.H. Yang, L.C. Ling, N. Carbon

Mater. 22 (2007) 153.