Preparation of microporous activated carbon and its electrochemical performance for electric double layer capacitor

ARTICLE IN PRESS Journal of Physics and Chemistry of Solids 70 (2009) 738–744

Contents lists available at ScienceDirect

Journal of Physics and Chemistry of Solids journal homepage: www.elsevier.com/locate/jpcs

Preparation of microporous activated carbon and its electrochemical performance for electric double layer capacitor Xiaojun He a,, Jiangwei Lei a, Yejing Geng a, Xiaoyong Zhang a, Mingbo Wu b, Mingdong Zheng a a b

School of Chemistry and Chemical Engineering, Anhui University of Technology, Ma’anshan 243002, China State Key Laboratory of Heavy Oil Processing, Institute of Heavy Oil Processing, University of Petroleum, Dongying 257061, China

a r t i c l e in f o

a b s t r a c t

Article history: Received 11 September 2008 Received in revised form 26 November 2008 Accepted 7 March 2009

Microporous activated carbons were prepared by microwave heating petroleum coke with potassium hydroxide as activation agent. Microporous activated carbons were characterized by infrared spectroscopy, X-ray diffraction and nitrogen adsorption/desorption isotherms. Electrochemical properties of an electric double layer capacitor using microporous activated carbon as electrode materials were investigated by constant current charge-discharge and electrochemical impedance spectroscopic techniques. The results show that the specific surface area and the pore volume of microporous activated carbon increase with increasing activation time before the activation time reaches 37 min. The microporous volume totals 94.0% in the microporous activated carbons and the average pore diameter of microporous activated carbon is 2.00 nm. Microporous activated carbons prepared in the activation time of 31, 35 and 37 min are named as AC-31, AC-35 and AC-37, respectively. Compared with AC-27 electrode, the internal resistance for ions transferring in AC-31, AC-35 or AC-37 electrode is relatively small. The specific capacitance of AC-31 is the biggest among the microporous activated carbons, and it retains 279.6 F g1 maintaining 93.5% capacity after 200 recycling number. & 2009 Published by Elsevier Ltd.

Keywords: C. Infrared spectroscopy C. X-ray diffraction D. Electrochemical properties

1. Introduction Electric double layer capacitor (EDLC) is an increasingly popular energy storage device due to its energy and power density bridging the gap between batteries and classical electric capacitors [1]. One of the applications of EDLC is as a second power system in electric vehicles. This application requires EDLC to have a small energy loss and a high energy density to improve the efficiency of electric vehicles. The energy loss of EDLC is caused by the series resistance it holds while the energy density of EDLC is dependent on its specific capacitance. The series resistance and specific capacitance of EDLC both originate from the polarizable electrode that is made of porous carbon materials. Activated carbon (AC) is a feasible porous material for energy storage and electrochemical purposes because of its unique surface characteristics. AC can be prepared by chemical, physical or chemical–physical activation methods [2–4]. However, only the surface of pores for AC, which the ions can access, can contribute to the double layer capacitance. Therefore, it is important to prepare AC electrode materials with pores that are large enough for the electrolyte to access completely, but small enough to ensure a large surface area [5]. In this paper, we report a feasible

 Corresponding author.

E-mail address: [email protected] (X. He). 0022-3697/$ - see front matter & 2009 Published by Elsevier Ltd. doi:10.1016/j.jpcs.2009.03.001

method to prepare microporous activated carbons by microwave heating petroleum coke with potassium hydroxide (KOH) as the activation agent, of which microporous activated carbons have larger pores and surface area.

2. Experimental 2.1. Preparation and characterization of microporous activated carbon Microporous activated carbons were prepared in microwave equipment (LWMC-205, China) by microwave heating petroleum coke with KOH as an activation agent. Petroleum coke is sieved to 74 mm before use. The content of Mad, Ad and Vdaf in petroleum coke is 1.20, 0.21 and 12.49 wt% while the element content of Cd, Hd and Sd is 90.11%, 4.25% and 0.81%,respectively, of which Mad is the water content in petroleum coke on the air dry basis, Ad is the ash content in petroleum coke on the dry basis and Vdaf is the content of volatile matter in petroleum coke on dry and ash-free basis. The petroleum coke was impregnated into KOH solution with carbon/KOH mass ratio of 1:3 in a glass container. The mixture was dried in a vacuum oven at 80 1C to remove the water in it. The resultant mixture was then placed in a corundum crucible with a small pore connected to atmospheric air. The corundum crucible with a loaded sample was placed on a rotary

ARTICLE IN PRESS X. He et al. / Journal of Physics and Chemistry of Solids 70 (2009) 738–744

300 AC-37 250 Transmittance (a.u.)

plate in microwave equipment for heating the sample uniformly. The output power of the microwave equipment was 250 W. The crucible was surrounded by a cylindric Al2O3 fiber composite to maintain the temperature in the range of 700–900 1C. The activated product was washed with distilled water to remove the ions in the product and dried at 100 1C in a vacuum oven. The functional groups in the dried AC were characterized by infrared spectroscopy (IR, Spectrometer Spectrum One). The graphitic nature of AC was checked by X-ray diffraction (XRD, Philips X’Pert Pro Super, CuKa radiation). N2 adsorption/ desorption using BELSORP measuring instruments (BELSORPmini, Japan, Inc.) was carried out to investigate the pore properties of the samples. All the samples were degassed at 250 1C under nitrogen flow for about 4 h prior to measurement. The nitrogen adsorption/desorption data were recorded at the liquid nitrogen temperature (77 K). The specific surface area was calculated using the Brunauere Emmette Teller (BET) equation. The diameter distribution was determined by the Barrette Joynere Halenda (BJH) method and the micropore diameter distributions were determined by the micropore analysis method (MP method) and the t-plot method. Finally, pore volumes were estimated to be the liquid volume of adsorption (N2) at a relative pressure of 0.99.

739

AC-35 AC-31

200

AC-27

150 C-H C=O

100

C-O

C=O

C=O

C-H

C=C

O-H

50 1000

0

2000

3000

4000

Wavenumbers (cm-1) Fig. 1. IR spectra of microporous activated carbons prepared at different activation times.

2.2. Preparation and measurements of microporous activated carbon electrode

3. Results and discussion 3.1. Surface morphology characteristics and porous properties of microporous activated carbon Fig. 1 shows IR spectra of microporous activated carbons prepared in different activation times. The weak bands at 1396, 2850 and 2900 cm1 originate from –C–H stretching vibration, which suggests the existence of the –CH2 group. A relatively sharper peak from –O–H stretching vibration is observed at 3390–3440 cm1, which implies the existence of water in microporous activated carbons. The weak peaks appearing at 590–610, 1600–1625 and 2360–2370 cm1 can be assigned to –CQO stretching vibration [6,7]. The peak at 1400 cm1 may be associated with –O–H bending deformation in the carboxylic acid group. A small peak that appeared at 1040 cm1 can be associated to R–O–R functionalities. It can be seen from Fig. 1 that the peak intensity of –O–H, –CQO and –C–H becomes weak with the increase of activation time. The results prove that the structure of microporous activated carbon becomes stable due to the decomposition of

1200 Intensity (a.u.)

AC was used as the base material for the electrode. Carbon black (CB) with a specific surface area of 550 m2 g1 (Cabot-Corp, China) was added to the electrode as conducting material. A mixture of active material, carbon black and poly(tetrafluoroethylene) (PTFE) at a weight ratio of 89:5:6 was pressed onto a foam nickel (current collector, 1.5  1.5 cm) followed by pressing the composite in an extrusion machine. Finally, the electrode (ca. 0.6 mm high) was dried at 120 1C for 2 h in a vacuum oven before use. Cyclic voltagrammetry and electrical impedance spectroscopy (EIS) analysis were carried out on an electrochemical workstation (CHI-760C, China) connected to a two-electrode test cell in a 6 M KOH solution. A pair of electrodes was set up to provide both an anode and a cathode using a symmetrical cell configuration. The capacitive behavior of the composite was investigated by constant current discharge in a land system (LAND, CT-2001A).

002

800

AC-37

400

AC-35 AC-31

0

AC-27 10

20

30

40

50

60

70

80

2θ (°) Fig. 2. XRD patterns of microporous activated carbons prepared at different activation times.

functional groups in microporous activated carbons with increasing activation time [8]. Fig. 2 presents XRD patterns of microporous activated carbons prepared at different activation times. Carbonaceous materials were determined by 2y at around 261 and 431. The crystallites of anisotropic AC would have larger size and good tropism if 0 0 2 peaks show narrow and sharp shape. On the other hand, isotropic pitch with broader 0 0 2 peaks would have more microstructural defects and noncrystallite carbon atoms [9]. It can be found from Fig. 2 that microporous activated carbons prepared at different activation times have more microstructural defects and noncrystallite carbon atoms, which is ascribed to the etching of KOH by microwave heating. The 2y at around 261 passes through a maximum with increasing activation time. The 2y is in the order: AC-27o AC-31o AC-35, which illustrates that the graphite degree of microporous activated carbon decreases with increasing activation time. According to the Bragg formula, the crystal interlayer distance (d002) will be decreased if y002 is increased. The d002 can be determined by 2d002 sin y ¼ l

(1)

ARTICLE IN PRESS X. He et al. / Journal of Physics and Chemistry of Solids 70 (2009) 738–744

600

Va (cm3 (STP) g-1)

Va (cm3 (STP) g-1)

500 400 300 200

550

450

desorption adsorption

400

BET (m2 g-1)

400

0 27

30 33 Activation time (min)

36

Fig. 4. Relationship between activation time and BET of microporous activated carbon.

0.4 AC-31 0.3

0.2

0.1

0.0 0

1

2 3 4 Pore diameter (nm)

5

6

Fig. 5. Relationship between pore volume and pore diameter of AC-31.

0.8

0.6 micropore+mesopore micropore

0.4

mesopore 0.2

28

30 32 34 Activation time (min)

36

38

Fig. 6. Effect of activation time on pore volume of microporous activated carbon.

350 0.0 0.2 0.4 0.6 0.8 1.0 Relative pressure

AC-31 AC-35

0 0.2

800

26

AC-35

100

0.0

1200

0.0

500

AC-37

AC-27

1600

VP (cm3 g-1)

where d002 is the interlayer distance of plane, y is the half-width value of reflection peak (0 0 2) and l is the wavelength of CuKa radiation. The pore formation in microporous activated carbon is due to the existing defects of graphite-like crystal layers. Microporous activated carbons are made up of graphite-like crystals. Therefore, if d002 is decreased, the pore size must be shrunk [10]. Fig. 2 shows that the y002 passes through a maximum with increasing activation time. So, the pore size in microporous activated carbons should have a minimum with increasing activation time. The yield of microporous activated carbon decreases from 93.9% to 60.2% when the activation time increases from 23 to 35 min. But the yield of AC does not show obvious change when the activation time reaches 37 min. K might be the activation product according to the furious reactions between K and H2O when the activation product from the corundum crucible is placed in water. The formed K2CO3 in the activation product can be detected by Ca(OH)2 because the solution of clear limewater becomes feculent when Ca(OH)2 is added into the clear solution of the activation product. Moreover, gas is formed when HCl solution is added into the deposit in feculent limewater. So, we know that K2CO3 is produced during the activation process. Fig. 3 shows the adsorption isotherms of AC-27, AC-31, AC-35 and AC-37. In all cases, the results correspond to Type I isotherms (characteristic of microporous materials) according to the IUPAC classification, as shown in the inset in Fig. 3. It is remarkable to note that the longer the activation time, the larger the pore volume. While relative pressure (P/P0) was about 0.2, the N2 adsorption was saturated. While P/P040.2, N2 began to be adsorbed onto the large pores or on the external surface, but the adsorbed amount was very small. Fig. 4 presents the effect of activation time on the specific surface area of microporous activated carbon. It can be seen from Fig. 4 that the specific surface area of AC-27 is only 869 m2 g1. The specific surface area increases fast with increasing activation time. The specific surface area of AC-31 reaches 1590 m2 g1 when activation time is 31 min, an increase of 83%. Then, the increase trend becomes slow and the specific surface area of AC-37 becomes 1760 m2 g1. The pore size distribution of AC-31 is shown in Fig. 5 as an example. It can be seen from Fig. 5 that the micropore is the main pore, of which the main pore ranges between 0.7 and 1.6 nm as well as 2.4 nm. The results of average pore size illustrate that the average pore size of AC-27, AC-31 and AC-35 is 2.00, 1.97 and 1.95 nm, respectively, while the average pore size of AC-37 is 2.00 nm. That is to say, the average pore size of microporous

Pore volume (cm3 g-1)

740

0.4 0.6 Relative pressure

0.8

1.0

Fig. 3. Adsorption isotherms of microporous activated carbons prepared at different activation times.

activated carbon shows a minimum with increasing activation time, which is consistent with the results of XRD analysis. The effect of activation time on the pore volume of microporous activated carbon is shown in Fig. 6. It can be seen from Fig. 6 that the total pore volume of micropore and mesopore increases with increasing activation time. The pore volume of AC-37 increases a little compared with the AC-31, which has the same changing trend with its specific surface area. The micropore

ARTICLE IN PRESS X. He et al. / Journal of Physics and Chemistry of Solids 70 (2009) 738–744

volume totals 94.0% in all pore volume while the mesopore volumes is only 4.0%. Fig. 7 illustrates the effect of activation time on the main pore volume of microporous activated carbon. It can be seen from Figs. 7(a) and (c) that the pore volume of bigger pore size (0.8, 0.9 and 2.4 nm) increases with increasing activation time while the pore volume of smaller pore size (0.6 and 0.7 nm) has a decreasing trend. It can be seen from Figs. 7(b) and (c) that the pore volume

Va (cm3 g-1)

0.4 0.6 nm 0.7 nm 0.8 nm 0.9 nm

0.3

0.2

741

of 1.0 and 1.1 nm is the main pore volume between 1.0 and 1.8 nm and passes through a maximum in an activation time of 35 min. On the contrary, the pore volume of 1.5 nm decreases when the activation time increases from 31 to 35 min. 3.2. Electrochemical characterizations 3.2.1. Cyclic voltammetry behavior of microporous activated carbon electrode Fig. 8 presents cyclic voltammograms (CVs) of the AC-31 electrode at different scan rates. It can be found that the current of AC-31 electrode increases with the increase of the scan rate. The current of the AC-31 electrode has a peak at around 0.49 V at 50 mV s1 scan rate because of the Faradaic current due to oxidation–reduction reaction of functional group containing oxygen. The curve shape becomes quasi-rectangular at higher scan rate, which is due to the resistance of ion migration in micropores because the increase of the scan rate aggravates the delay of the current to reach a horizontal value after reversal of the potential scan.

0.1

0.0 26

28

36

38

36

38

1.1 nm 1.2 nm 1.3 nm 1.5 nm 1.6 nm 1.7 nm 1.8 nm

0.020

0.016 Va (m3 g-1)

30 32 34 Activation time (min)

0.012

0.008

0.004 26

28

30 32 34 Activation time (min)

3.2.2. Charge-discharge behavior of microporous activated carbon electrode Chronopotentiogram of AC-31 electrode at different chargedischarge current is shown in Fig. 9. It can be seen from Fig. 9 that there is a sudden drop of potential at the beginning stage of discharge, which is associated with the ohmic resistance of the cell. The equilibrium series resistance (ESR) of the capacitor was obtained from the discharge curves [11]. The specific capacitance (Csp) and ESR of AC-31 and AC-35 electrodes are shown in Fig. 10. The Csp was calculated from the discharge slope (11 cycles) [12]. It can be observed from Fig. 10(a) that the average Csp of the AC-31 electrode decreases from 298.8 to 279.6 F g1 at the discharge density of 100 mA g1 before 95 recycling number, retaining 93.5% capacity. Then, the average Csp of AC-31 electrode almost retains 279.6 F g1. The Csp of AC-35 electrode decreases from 275.3 to 258.8 F g1 before 95 recycling number, maintaining 94.0% capacity. The ESR of the AC-31 electrode increases from 2.32 to 3.73 O while that of the AC-35 electrode increases slowly from 1.13 to 1.36 O. So, the Csp of AC-31 electrode is greater than that of AC-35 electrode while the ESR of the former is also bigger than that of the latter. AC-31 and AC-35 with lower specific surface area and total pore volume showed bigger specific capacitance value as 258.8–298 F g1 at the discharge density of 100 mA g1, which are

0.3

0.075

AC-31

25 mV s-1

1.0 nm 2.4 nm

0.2

10 mV s-1 5 mV s-1

Current (A)

0.060 Va (cm3 g-1)

50 mV s-1

0.045

0.1

2 mV s-1

0.0 -0.1

0.030 -0.2

26

28

30 32 34 Activation time (min)

36

38

Fig. 7. Effect of activation time on pore volume of microporous activated carbon with different pore sizes.

-0.3 -1.0

-0.5

0.0 Potential (V)

0.5

Fig. 8. CVs of AC-31 electrode at different scan rates.

1.0

ARTICLE IN PRESS 742

X. He et al. / Journal of Physics and Chemistry of Solids 70 (2009) 738–744

1.0 AC-31 0.8 100 Voltage (V)

Table 1 Specific capacitance and ERS comparison of AC-27, AC-31 and AC-35 electrode.

200 mA g-1

Discharge current (mA g1)

mA g-1

20 mA g-1

0.6

20 50 100 200 300

0.4

0.2

0.0 0

2000

4000 Time (s)

6000

8000

Fig. 9. Charge-discharge curves of AC-31 electrode at different current densities.

10

250 200

6

average ESR simulated curve of ESR

150

4 100 2

50

0

20

40

60

80 100 120 140 160 180 200 Recycling number

8

300 250

average CSP

6 ESR (Ohm)

simulated cuve of CSP

200 150

4 average ESR simulated curve of ESR

100

2 50 0

0 0

50

100 150 Recycling number

200

Fig. 10. Cyclic performance comparison of the two electrodes: (a) AC-31 and (b) AC-35.

higher than the recently reported results (o257.0 F g1) [3]. The above difference might be ascribed to the bigger average pore size of petroleum coke-based AC by microwave heating activation because the average pore size of petroleum coke-based AC-31 and AC-35 is in the range of 1.95–1.97 nm while the average pore size of corn grain-based AC ranges between 1.03 and 1.18 nm [3].

AC-31

AC-35

Csp (F g1)

ESR (O )

Csp (F g1)

ESR (O)

Csp (F g1)

ESR (O)

300 282 275 245 230

23.7 16.2 10.0 7.4 5.5

330 316 309 303 300

7.1 4.1 2.8 2.2 2.0

302 290 285 274 266

3.4 1.8 1.3 1.3 1.2

The Csp and ESR of microporous activated carbon electrode at different discharge current according to the results of the third charge-discharge are presented in Table 1. It can be seen from Table 1 that Csp passes through a maximum with the increase of activation time from 27 to 35 min. The Csp and ESR of AC-31 electrode in Table 1 decrease with increasing discharge current, of which the decreasing capacitance with increasing discharge current density implies the potential difference in micropores because of the ohmic resistance of the electrolyte [3]. The ESR of microporous activated carbon electrode decreases with the increase of activation time for AC-27, AC-31 and AC-35 electrodes. Ragone plots for AC-31 and AC-35 electrodes about the dependence between power density and energy density are shown in Fig. 11, which are quite useful to determine the application potentials for different microporous activated carbons. The energy (E) was calculated as [13] E¼

1 CV 2 7:2

(2)

where C is the capacitance of the two-electrode capacitor and V is the voltage decrease in discharge. The power (P) was calculated as Eq. (3) P ¼ VI

0

0

Specific capacitance (Fg-1)

ESR (Ohm)

8

Specific capacitance (Fg-1)

300 average CSP simulated cuve of CSP

AC-27

(3)

where V is the voltage decrease and I is the discharge current in discharge process. The energy density and power density were calculated by dividing E and P by the mass of porous carbons in the capacitor cell. As illustrated by Fig. 11, the energy density of AC-35 electrode increases from 15.7 to 20.4 Wh kg1 while the energy density of AC-31 electrode increases from 14.5 to 21.2 Wh kg1 when the power density of AC-31 or AC-35 electrode decreases from 0.25 or 0.28 to 0.02 kW kg1. Comparatively, AC-31 electrode shows broader energy density range than AC-35 electrode. This can be ascribed to its larger content of micropore between 0.6 and 0.7 nm (o1 nm), which favors the fast transport of solvated ions during the chargedischarge process [14].

3.2.3. Impedance spectrum of microporous activated carbon electrode Fig. 12 shows Nyquist curves of the four microporous activated carbon electrodes. It can be observed from Fig. 12 that AC-31 has an ideal Nyquist curve than other samples with less contact resistance and diffusion resistance. At the higher frequency range, AC-31 has less contact value with the real axis and hemicycle than other samples (the inset in Fig. 12), which indicates smaller contact resistance and charge transfer resistance. At the lower frequency range, the curve of AC-31 is nearly vertical to the real axis, but AC-27 has the largest slope, which indicates that the diffusion resistance of electrolyte ions in the pore of AC-31 is smaller than that of AC-27. So, AC-31 has the better route for the ions to move.

ARTICLE IN PRESS X. He et al. / Journal of Physics and Chemistry of Solids 70 (2009) 738–744

0.30

Table 2 Simulation results of equivalent circuit elements from the simulation of the impedance data.

AC-31 Electrode samples

Rs (O)

W (O)

Rct (O)

Q (F)

n

w2

AC-27 AC-31 AC-35 AC-37

0.4419 0.5488 0.6369 0.4738

0.01339 0.09773 0.05333 0.01310

2.587 0.2246 0.3588 0.8817

0.3529 6.677 3.483 3.973

0.5289 0.9211 0.9099 0.8850

4.881  103 6.917  104 1.221  103 2.536  103

AC-35 0.20 0.15 0.10

Table 3 The error of equivalent circuit elements from the simulation of the impedance data.

0.05

Electrode samples

16

18

20

Energy density (Wh

22

kg-1)

AC-27 AC-31 AC-35 AC-37

Fig. 11. Ragone plots of AC-31 and AC-35 electrodes.

3.0

0.8 Image impedance (Ohm)

AC-27 AC-31 AC-35 AC-37

4.5

AC-27 AC-31 AC-35 AC-37

0.6

Error (%) Rs

W

Rct

Q

n

5.617 1.615 2.149 4.874

5.361 13.71 13.59 9.376

3.641 3.881 3.807 2.650

4.212 1.983 2.121 3.983

3.335 0.8888 0.9901 1.834

10

70 Model: R(WR)Q

Wgt: Modulus IZI, Msd. IZI, Calc.

0.4

0.0 0.5

1.0 1.5 Real impedance (Ohm)

2.0

40 Angle, Msd. Angle, Calc.

0.1

1.5

60 50

1

0.2

|Z| (Ohm)

0.00 14

30

Angle (deg)

Power density (kW kg-1)

0.25

Image impedance (Ohm)

743

20 10

0.0

0.01 0.01

0

2 Real impedance (Ohm)

4

Fig. 12. AC impedance spectrum of microporous activated carbon in 6 M KOH electrolyte.

W Rs

CPE Q

Rct Fig. 13. Equivalent circuit of AC electrode. [Rs: the contact resistance; Q: capacitance—a constant phase element (CPE); Rct: the charge transfer resistance and W: the Warburg resistance].

Several electrical circuits were tested by nonlinear leastsquares fitting of the experimental impedance data. The equivalent circuit shown in Fig. 13 was found to give excellent fits down to frequencies including the low-frequency vertical line corresponding to the low-frequency bulk capacitance of AC electrodes. The simulation results of the impedance data for different AC electrodes using ZSimpWin software are illustrated in Table 2 while the errors of the parameters for corresponding equivalent circuits from fitting EIS data of the AC electrode are listed in Table 3.

0.1

1 10 100 Frequency (Hz)

1000

0 10000

Fig. 14. |Z| and phase angle & frequency for AC-31 electrode: (K) measured and (&) calculated.

The average error (w2) of the fits in Table 2 for the different impedance spectra of the AC electrode is relatively smaller, ranging from 6.917  104 to 4.881 103. The error of the corresponding element is in the range of 0.8888–13.71% (shown in Table 3), which proves that the selected equivalent circuits are right for the investigated electrodes. Double-layer capacitance is expressed as constant phase element (Q) in Table 2. From a comparison of Rct from different AC electrodes, the ionic charge transfer resistance of AC-27 is the biggest in all samples. The results in Table 2 proved that the Rct of AC-31 is the smallest while its Q is the biggest among the mentioned AC electrodes. That is to say, the AC-31 electrode has the smallest charge transfer resistances and the biggest capacitance. The effect of frequency on the resistance (|Z|) and phase angle for AC-31 electrode is shown in Fig. 14. It can be seen from Fig. 14 that |Z| and phase angle of AC-31 electrode decrease with the increase of frequency, which is consistent with the recently reported results [15].

4. Conclusions Microporous activated carbons were prepared by microwave heating petroleum coke with KOH as the activation agent.

ARTICLE IN PRESS 744

X. He et al. / Journal of Physics and Chemistry of Solids 70 (2009) 738–744

The average pore diameter of microporous activated carbons is 2.00 nm and the microporous volume totals 94.0% in the microporous activated carbons. The specific surface area and the pore volume of microporous activated carbon increase with increasing activation time before the activation time reaches 37 min. The internal resistance for ions transferring in AC-31, AC-35 or AC-37 electrode is relatively small. The specific capacitance of AC-31 is the biggest among the microporous activated carbons, and it retains 279.6 F g1 maintaining 93.5% capacity after 200 recycling number.

Acknowledgements This work was supported by Natural Science Foundation of China (no. 50802002), Natural Science Research Foundation (no. KJ2008A120), Foundation for Youth Teacher in University of Anhui Province (no. 2008JQ1026ZD) and the Fund of Provincial Innovative Group for Processing & Clean Utilization of Coal Resource. References [1] R. Ko¨tz, M. Carlen, Principles and applications of electrochemical capacitors, Electrochim. Acta 45 (2000) 2483–2498. [2] Z.H. Hu, H.M. Guo, M.P. Srinivasan, Y. Ni, A simple method for developing mesoporosity in activated carbon, Sep. Purif. Technol. 31 (2003) 47–52. [3] M.S. Balathanigaimani, W. Shim, M. Lee, C. Kim, J. Lee, H. Moon, Highly porous electrodes from novel corn grains-based activated carbons for electrical double layer capacitors, Electrochem. Commun. 10 (2008) 868–871.

[4] Y.B. Ji, T.H. Li, L. Zhu, X.X. Wang, Q.L. Lin, Preparation of activated carbons by microwave heating KOH activation, Appl. Surf. Sci. 254 (2007) 506–512. [5] B. Xu, F. Wu, R.J. Chen, G.P. Cao, S. Chen, Z.M. Zhou, Y.S. Yang, Highly mesoporous and high surface area carbon: a high capacitance electrode material for EDLCs with various electrolytes, Electrochem. Commun. 10 (2008) 795–797. [6] S. Goyanes, G.R. Rubiolo, A. Salazar, A. Jimeno, M.A. Corcuera, I. Mondragon, Carboxylation treatment of multiwalled carbon nanotubes monitored by infrared and ultraviolet spectroscopies and scanning probe microscopy, Diamond Relat. Mater. 16 (2007) 412–417. [7] H.M. Zhu, J.H. Yan, X.G. Jiang, Y.E. Lai, K.F. Cen, Study on pyrolysis of typical medical waste materials by using TG-FTIR analysis, J. Hazard. Mater. 13 (2008) 670–676. [8] I.A.W. Tan, A.L. Ahmad, B.H. Hameed, Adsorption of basic dye using activated carbon prepared from oil palm shell: batch and fixed bed studies, Desalination 225 (2008) 13–28. [9] Y.S. Wang, C.Y. Wang, C.Y. Guo, Z.Q. Shi, Influence of carbon structure on performance of electrode material for electric double-layer capacitor, J. Phys. Chem. Solids 69 (2008) 16–22. [10] J.C. Xie, X.H. Wang, J.Y. Deng, L.X. Zhang, Pore size control of pitch-based activated carbon fibers by pyrolytic deposition of propylene, Appl. Surf. Sci. 250 (2005) 152–160. [11] X.J. He, L. Jiang, S.C. Yan, J.W. Lei, M.D. Zheng, H.F. Shui, Direct synthesis of porous carbon nanotubes and its performance as conducting material of supercapacitor electrode, Diamond Relat. Mater. 17 (2008) 993–998. [12] Y. Nian, H. Teng, Nitric acid modification of activated carbon electrodes for improvement of electrochemical capacitance, J. Electrochem. Soc. 149 (2002) A1008–A1014. [13] B. Fang, Y.Z. Wei, M. Kumagai, Modified carbon materials for high-rate EDLCs application, J. Power Sources 155 (2006) 487–491. [14] J. Chmiola, G. Yushin, Y. Gogotsi, C. Portet, P. Simon, P.L. Taberna, Anomalous increase in carbon capacitance at pore sizes less than 1 nm, Science 313 (2006) 1760–1763. [15] M. Sipahi, E.A. Parlak, A. Gul, E. Ekinci, M.F. Yardim, A.S. Sarac, Electrochemical impedance study of polyaniline electrocoated porous carbon foam, Prog. Org. Coat. 62 (2008) 96–104.