ECHE789B Special Project – Lingyun Liu

05/28/2002

ECHE789B Special Project

Experimental Studies of Asymmetric Capacitors

Instructor: Dr. Popov By: LINGYUN LIU May 5, 2002

1/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

ABSTRACT The performance of nickel hydroxide asymmetric capacitors is studied and compared with that of carbon symmetric capacitors. The asymmetric capacitor has better energy efficiency. Optimize the design parameters of the asymmetric capacitors such as thickness, state-of-charge. The behavior of the devices under different discharging rates will also be studied. Ragone plots are used to evaluate the power density and achievable energy density for both symmetric and asymmetric capacitors. The results show some unexpected tendency. The possible reasons are analyzed.

2/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

INTRODUCTION Electrical energy storage is needed in lots of area. The energy storage and conversion devices include batteries, fuel cells, capacitors etc. Capacitor is one kind of the mostly common used energy storage backup devices. Capacitors store energy by charge separation. It can provide pulse electrical energy. Generally, there are three kinds of capacitors, say, film capacitors, electrolytic capacitors, and electrochemical capacitors. Electrochemical capacitor is also called supercapacitor or ultracapacitor. Electrochemical capacitors may improve battery performance in terms of power density or may improve capacitor performance in terms of energy density when combined with the respective device. At the same time, electrochemical capacitors are expected to have much longer cycle life. Figure 1 and figure 2 show the reason why electrochemical capacitors are concerned. They fill in the gap between batteries and converntional capacitors such as electrolytic capacitors or metallized film capacitors. In terms of specific energy as well as in terms of specific power this gap covers several orders of magnitude [1].

Fig. 1. Sketch of Ragone plot for various energy storage and vonversion devices. The indicated areas are rough guide line (by R. Kotz and M. Carlen)

3/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

Fig. 2. Ragone plane: available energy of an energy storage device for fixed power. Different types of energy storage devices are typically located in different regions. Characteristic times correspond to lines with unity slope. Every energy storage device is represented by a curve E(P) (inset). Internal dissipation and leakage losses lead to a drop of the energy for sufficiently high and low power (by T. Christen, M.W. Carlen)

The electrochemical capacitor is constructed like a battery in that it has two electrodes immersed in an electrolyte with a separator between the electrodes [2]. There are two types of electrochemical capacitors are highly concerned nowadays, that is, double-layer capacitors and hybrid capacitors using pseudocapacitance. In this paper, we call the former one as symmetric capacitors since the properties are same in both positive electrode and negative electrode, and the latter one as asymmetric capacitors since the capacitor are fabricated with double-layer capacitance material ( i.e. carbon) as negative electrode and pseudocapacitance material (i.e. metal oxides) as the positive electrode. Energy is stored in the double-layer capacitor as charge separation in the double-layer formed at the interface between the solid electrode material surface and the liquid electrolyte in the micropores of the electrodes. The mechanism is shown in figure 3. The capacitance depends on the characteristics of the electrode material (i.e. surface area and pore size distribution) [2].

4/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

Fig. 3. Schematic of a double-layer capacitor. (by A. Burke)

Unlike the double-layer capacitor which is nonfaradic, pseudocapacitor use faradic process to store energy. The electrical energy is converted to chemical energy through charge and the chemical energy release as electrical energy when discharge. Metal oxide materials are usually used to obtain pseudocapacitance. Lots of researchers are working on it and trying to find higher efficiency and lower cost electrode material. There are great achievements on this. RuO2 is one of the most recommended materials since its high capacitance and energy density [3], but the material is quite expensive which limits the widely application. Other materials such as CoO, V2O5 have the same problem [4]. Compared with these materials, nickel oxides material has apparently advantages: 1. low cost, 2. low toxicity, and 3. much knowledge of electrochemical characteristics of nickel oxides (hydroxides) can be obtained from nickel batteries’ study. The asymmetric capacitors use Ni(OH)2/Co(OH)2 thin films as the positive electrode and traditional porous carbon electrodes as the negative. The pseudocapacitance of the nickel hydroxide comes from the reaction taken place when charge and discharge, as following:

5/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

 → Ni(OH) 2 + OH − NiOOH + H 2O + e − ←  disch arg e ch arg e

Figure 4 illustrates the capacitance mechanism of nickel hydroxides film.

Metal Substrate eNiOOH/Ni(OH)2

echarge

discharge

H+

H+ Electrolyte H2O OH-

H2O

OH-

Fig. 4. The mechanism of pseudocapacitance of Ni(OH)2/NiOOH

The former experimental data verify that the potential of Ni(OH)2/Co(OH)2 thin films would maintain high level during a long time when discharge because of reaction. It is hypothesized that the energy density of this device can be a factor of four times larger than traditional electrochemical capacitor due to the extremely high capacity of nickel hydroxide. The data indicates that the capacity of nickel hydroxide is about 10 times than that of carbon even though the mass of the nickel hydroxide is 1% the mass of the carbon. Therefore, it is hypothesized that the voltage of the positive electrode will remain essentially constant during the discharge of a device yet contributes negligible weight. This hypothesis will be fully tested. The mathematical models of both asymmetric capacitors and symmetric capacitors predict the performance and energy efficiency of the capacitors. Figure 5 shows the potential profile through electrodes when discharge. The area covered by asymmetric capacitors is almost twice as that of symmetric capacitors, which indicate the asymmetric capacitor has much lager capacitance. Figure 6 is the prediction of the energy density and power density. Both of them are several magnitudes large for asymmetric capacitors than for symmetric capacitors.

6/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

EXPERIMENTAL A computer-controlled EG&G Princeton Applied Research M273 Potentiostat/ galvanostat using the M270 software is used in studies. Thin films of Ni(OH)2/Co(OH)2 have been fabricated by depositing them on nickel foil film with exposed area of 1cm2. The deposition is taken at 56°C (RM6 Lauda, Brikmann) in the beaker containing 1.8M Ni(NO3)2, 0.18M Co(NO3)2, and 0.075M NaNO3 in the solvent of 50 vol% ethanol. A cathodic current density of 5.0mA/cm2 was applied for 25min which according to these previous deposition studies should result in 350µg films with a capacity of 277mC (i.e. 790 C/g). The expected capacity was confirmed by performing cyclic voltammetry on nickel hydroxide in 3wt% KOH and integrating the area under the reduction peak of a stable cyclic voltammogram [5]. Before study in KOH solution, the nickel hydroxide film needs to be rinsed in DI water. Saturated Calomel electrode (SCE) and platinum mesh are used as reference and counter electrode. The film is constant current charged at first with the current density of 1mA, and then cycles in solution for 10 cycles, and then steady state capacitance is measured. After that, the constant current discharge process is taken with the current density of 0.02mA. XC-72 is used as negative electrode material. The carbon has been dispersed in 5 wt% Nafion solution, and small volume of isopropanol has been added to enhance the dispersion. The mixture is stirred for more than 8 hours to get good ink. Spray the ink on decals with the area of 10cm2 with the sprayer (P-163) Pssache Millennium Set). The decals had been cleaned and weighted before use. Dry the sprayed decals at the temperature of 105˚C for 10 min, then weight it after cool. The average carbon loading can be calculated. Porous carbon electrode has been fabricated by press the decal into Nafion 117 membrane (Dupon) at 150˚C. Pelt the decal off after cool. Hot press (Carver, Inc.) has been used. Prepare several carbon electrodes with different average carbon loadings for use. To fabricate asymmetric capacitor, punch 1cm2 carbon electrode–membrane assembly as negative electrode, and the nickel hydroxide film prepared before as positive electrode. On the other hand, carbon electrode-membrane-carbon electrode assembly severs as symmetric capacitor. The sketches of the capacitors are shown in Figure 6.

7/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

Fig. 6. The rough sketches of the capacitors. (a) asymmetric capacitor, (b) symmetric capacitor.

T-cells have been use to handle the test with carbon electrode as working electrode and SCE as reference electrode. Both of these capacitors have been cycled over potential range from 0mV to -600mV for at least 20 cycles. 3 wt% KOH solution is used as electrolyte. Different currents will be used to discharge the capacitors. The capacity will be measured for various discharging rate.

8/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

RESULTS AND DISCUSSION Nickel hydroxide film study — Figure 7 shows the CV of nickel hydroxide film at a sweep rate of 5mV/s. The redox peaks represent the oxidation/ reduction reaction of charge and discharge. The ideally reversibility predict the long cycle life. Figure 8 is the constant current charge and discharge curve of nickel hydroxide film. There are two plateaus in charge curve. The first one is the oxidation of Ni(OH)2 with the formation of NiOOH, and the second one is the water dissociation with evolution of oxygen. In discharge curve, the potential drop sharply at the initial and end regions. Within the large region between then, the potential change very slowly. If initial state-of-charge (SOC) is set around 50%, the potential is almost stable when charge or discharge 10mV. That’s the potential window of operation. 10 8 6 I, mA

4 2 0 -2 -4 -6 -8 0

100

200

300

400

500

E, mV

Fig. 7. CV of nickel hydroxide film with the scan rate of 5mV/s

9/16

600

ECHE789B Special Project – Lingyun Liu

05/28/2002

Potential, mV vs. SCE

500 400 300 200 100 0 0

500

1000

1500

2000

2500

time, s Fig. 8. Constant charge and discharge curve of the nickel hydroxide film. Charge rate is 1mA and discharge rate is 0.02mA

Asymmetric capacitors and symmetric capacitors — Cyclic voltammetry is used to evaluate the capacitance. Figure 9 is the cyclic voltammograms for asymmetric capacitors as well as symmetric capacitor. Boxed shape curve shows the characteristics of capacitor behavior, that is, the current remain constant when cell potential changes. From the area of the steady state cyclic voltammogram, the capacity of capacitors is obtain. Figure 10 illustration the difference between those capacitors. The capacity of asymmetric capacitor is bigger than that of symmetric capacitor. For asymmetric capacitors with different carbon loadings, the capacity increases with respect to carbon loading at first, then decreases. The reason of this phenomenon maybe is because when carbon layer is too thick only surface carbon contact thoroughly with electrolyte. The lower carbon cannot provide double-layer capacitance, but extremely increases the resistance.

10/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

400 300 200 i, mA

100 0 -100-600

-500

-400

-300

-200

-100

0

1.8mgC/cm^2

-200

2.8mgC/cm^2

-300

5.1mgC/cm^2

-400

6.2mgC/cm^2

-500

C-C 0.5mg/cm^2

E, mV Fig. 9. CVs of symmetric capacitor and asymmetric capacitors with different average carbon loadings. 30 25 20

Carbon Loading, mg/cm^2

15

Capacity, mC

10 5

EC as

#5

as

ym

EC ym

EC #4

#3

as

ym

EC ym

#2

as

ym as #1

C -C

sy

m

EC

EC

0

Fig. 10. Capacity comparison for different type and different carbon loading capacitors.

Constant current charge and discharge are used to study the capacitor performance and energy behavior. Figure 11 is the discharge curve for different carbon loading capacitors.

11/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

Its behavior shows the same trend as we got from CVs. Figure 12 is the discharge curve of constant carbon loading with different current. With the current decreasing, the discharge time increasing. This do make sense. Constant discharge curve (1.8mg/cm^2 carbon loading)

400 10mA/s

300

1mA/s

E, mV

0.1mA/s 0.01mA/s

200

0.001mA/s

100 0 0

500

t. s

1000

1500

Fig. 11. Constant current (1mA) discharge for different carbon loading capacitors. 400

E, mV

350 300

0.5mg/cm^2

250

1.8 2.8

200

3.16

150

5

100

6

50 0 0

2

4

6

8

10

12

t, s

Fig. 12. Constant current discharge for asymmetric capacitors with different current.

Specific discharge capacitance is calculated as: C = (2 × I × t) / (w × ∆E)

12/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

where C is the specific discharge capacitance (F/g), I the discharge current (A), t the cutoff time (s), w the weight of the capacitor (g), and ∆E the potential difference (V). Figure 13 shows the specific discharge capacitances for different capacitors. 1.8 1.6 capacitance F/g

1.4 1.2

1mA/s

1

0.1mA/s

0.8

0.01mA/s

0.6

0.001mA/s

0.4 0.2 0 0

1

2

3

4

5

6

7

Carbon loading, mg/cm^2

Fig. 13. Specific discharge capacitance.

The energy density and power density are also studied based on the constant current discharge curve. Ragone plots was developed to compare and contrast those capacitors. The energy density (W/kg) calculate from: E=

i τc ∫ Vdτ w0

and power density (W-h/kg) from:

P=

i τc ∫ Vdτ wτ c 0

where τ c is the cutoff time (s)[7]. The Ragone plots are illustrated in figure 14. At low current region, the curves appear unusually push-back in stead of straight lines with almost constant energy density. There are two possible reasons. First, maybe because the limiting factor is not carbon loading as we expected. That means the nickel hydroxide film is not enough. From the former study, it is sure that the nickel hydroxide film have much bigger capacity than that of the carbon, but this not means this reason is not possible. After a bunch of operations, some Ni(OH)2 pelt from the nickel foil. Perhaps the

13/16

ECHE789B Special Project – Lingyun Liu

05/28/2002

substance left is not enough to make the carbon as the limitation. This can be eliminated through increasing nickel hydroxide film loading and enhance nickel foil surface characteristics. Another probable reason is the self-discharge. The nickel hydroxide film is prepared from nickel nitrate. The nitrate ion is known as extremely easy to selfdischarge. To avoid this possibility, the nickel hydroxide film should be rinse more thoroughly or find some better way to get rid of nitrate ion.

100

Cut-off potential: 0V

Power Density, W/kg

10

1.8mgC/cm^2

1

2.8mgC/cm^2 0.5mgC/cm^2 5.1mgC/cm^2

0.1

6.2mgC/cm^2 0.01

0.001 0.0001

0.001

0.01

Energy Density, W-h/kg

Fig. 14. Ragone plots for asymmetric capacitors.

14/16

0.1

ECHE789B Special Project – Lingyun Liu

05/28/2002

CONCLUSIONS Nikel hydroxide has big capacitance which can be used in hybrid capacitor. Asymmetric capacitor is superior to symmetric capacitor. The average carbon of the negative electrode in asymmetric capacitor heavily affects capacitor behavior. The critical carbon loading may exist. Most experiments should be done in future work to confirm the results and eliminate the factors discussed.

15/16

ECHE789B Special Project – Lingyun Liu

REFERENCES [1] R. Kotz, M. Carlen, Electrochimica Acta 45 (2000) 2484. [2] Andrew Burke, J. Power Sources 91 (2000) 39. [3] J.P. Zheng, T.R. Jow, J. Power Source 62(1996) 155. [4] S. Chin, S. Pang et al., J. Electrochem. Soc. 149(200) A379. [5] V. Srinivasan, J. Weidner, J. Electrochem. Soc. 147(2000) 880. [6] R. Huggins, Solid State Ionics 134(2000) 179 [7] V. Srinivasan, J. Weidner, J. Electrochem. Soc. 146(1999) 1654. [8] J. Weidner, P. Timmerman, J. Electrochem. Soc. 141(1994) 346. [9] W.G. Pell, B.E. Conway, J. Power Source 63(1996) 258 [10] G. Amatucci, F. Badway et al.. J. Electrochem. Soc. 148(2001) A930

16/16

05/28/2002