Alkaline poly(ethylene oxide) solid polymer electrolytes. Application to nickel secondary batteries

EIecrroekimicn

Acra, Vol. 40, No. 13 14, pp. 2449.2453. 1995 Copyright 0 1995 Elsevier Science Ltd. Printed in Great Britain. All rights reserved 0013~4686/95 $9.50 + 0.00

0013-4686(95)00212-x

ALKALINE POLY(ETHYLENE OXIDE) SOLID POLYMER ELECTROLYTES. APPLICATION TO NICKEL SECONDARY BATTERIES J. F. FAWARQUE~,S. GUINOT,N. Bouzw, E. SALMONand J. F. PENNEAU* Laboratoire d’Electrochimie Industrielle, CNAM-2, rue ContG-75003 Paris, France * EDF-Dkpartement des Etudes et de la Recherche, Les Renardihres-77250 Moret-sur-Loing, France 6 April 1995)

(Received

Abstract-Water-containing and anhydrous alkaline PEO-based solid polymer electrolytes have been prepared, involving two different technologies. They exhibit a high ionic conductivity, and values up to 10e3 Scm-’ are obtained for some compositions at room temperature. Conductivity studies and thermal analysis are presented, and the conductivity vs. temperature behaviour of each type of solid polymer electrolyte studied is discussed. Preliminary results are reported concerning the cyclability of cells involving a nickel positive electrode, and a cadmium or a zinc negative electrode.

Key words: alkaline batteries.

solid polymer

electrolyte,

poly(ethylene

INTRODUCTION

should be addressed.

nickel secondary

EXPERIMENTAL

Solid polymer electrolytes are ionic conductive solutions of salts in a polymeric solvent, typically poly(ethylene oxide) (PEO). The study of such new electrolytes was initiated by Wright et aI.[l] who carried out the first conductivity measurements of polymer-salt complexes, and concretised by Armand et aI.[2-31, who was the first to recognise the real practical potential use of PEO-based electrolytes. The diversity and the complexity of the current research in the field of polymer electrolytes can be explained by their attractive advantages as compared to conventional solid or liquid electrolytes. For example, problems like chemical and electrochemical instability, corrosion and gas emission, can be avoided in such polymer-based systems, which can in addition be easily processed due to their plasticity. This undoubtedly makes solid polymer electrolytes real good candidates for applications like batteries and other electrochemical devices[4]. Among numerous systems, lithium-based solid polymer electrolytes have been the most extensively studied materialsC3, 51. We present here a new family of solid polymer electrolytes, whose specificity lies in their composition, based on PEO and potassium hydroxide (KOH). Several compositions have been prepared, and two have been mainly studied, an anhydrous one and a water-containing one. The thermal and conductivity properties of those two compositions have been investigated. As an application, a prospective study of the cyclability of nickel secondary batteries, namely nickel/cadmium and nickel/zinc, is reported here.

t To whom correspondence

oxide), ionic conductivity,

Commercial PEO 20.000 (Fluka av.mol. wt. 2 x 104), KOH (Fluka), and methanol (Merck) were used as starting materials for solid polymer electrolytes preparation. The water-containing composition was prepared by direct mixing of PEO, KOH and H,O, at approximately 90°C. The anhydrous composition was prepared by the conventional solvent casting method. Equivalent amounts of PEO and KOH were dissolved in methanol, and were mechanically stirred for several hours at room temperature, so that the solvent was allowed to evaporate slowly. After casting on an appropriate support, the last traces of solvent were removed under vacuum. The ionic conductivity of both compositions was determined by using the UCcomplex impedance method. A typical button cell composed of two expanded nickel grid electrodes was used for the measurements, which were carried out over the frequency range of 0.1 Hz to 2 x lo6 Hz using a Schlumbezger 1250 Frequency Response Analyser and 1286 Electrochemical Interface controlled by a Compaq microcomputer. All analysed cells were previously kept at 80°C for 20min and then allowed to cool at room temperature, to ensure good contact between the electrolyte and the electrodes. Two or even three series of measurements were performed for each of these two complexes in different button cells, and the samples were heated and cooled five times and data collected on heating. With an acceptable reproducibility, the fifth heating cycle data are reported here. Fifth heating-cooling cycle measurements at a frequency value of 104Hz are also presented. All impedance measurements were made after half an hour temperature stabilisation. The total ionic conductivity was calculated using equation (1), where 1 and A are the thickness of the electrolyte and the

2449

J. F. FAUVARQUE et al.

2450

area of the electrodes respectively, and R, is the resistance of the electrolyte estimated from UCimpedance data. a(Scm-‘)

=

l(cm) R&2) x A(cm2)

(1)

A SETARAM 101 differential scanning calorimeter (DSC) was used to determine the melting temperature and other thermal characteristics of the two compositions, using stainless steel sealed pans. The samples were heated at a rate of 3”Cmin- ‘, over a temperature range of 0-100°C in nitrogen atmosphere. Each sample was run three times over the temperature range previously indicated. The temperatures reported are onset temperatures. RESULTS AND DISCUSSION

Water-containing composition of 60% wt.% PEO, 30wt.% KOH and lOwt.% H,O The Arrhenius plot of the total ionic conductivity is shown in Fig. la, exhibiting an unusual behaviour of the electrolyte. Indeed, three domains can be -2.5 , a

l

l* l* .

.

.

. .

2.7

.

.

3.1

2.9

3.3

3.5

(I’XOfO/K-I

observed on the curve log (a) vs. f (l/T), in the overall temperature range. Two domains, from room temperature to 60°C and from 65°C to the maximum temperature, show a linear increase of the conductivity with increasing temperature, which is the normal situation expected for a PEO-based solid polymer electrolyte. A third domain, in the temperature range of 60-65”C, surprisingly shows a very important decrease of the conductivity. This transition is not the usual one accounted for PEO-based solid polymer electrolytes, such as Arrhenius type. Figure lb shows typical impedance plots for the electrolyte, at three different temperatures from the three different domains. DSC curves of this water-containing composition are shown in Fig. 2, exhibiting two endothermic peaks. The one at 66°C could be attributed to the melting, either of the crystalline phase of PEO or of another crystalline phase existing in the electrolyte. The other at 36°C has not been clearly identified yet. Further investigations such as X-ray analysis are under progress. The unusual conductivity vs. temperature behaviour of the 6Owt.% PEO, 30wt.% KOH and lOwt.% H,O composition has also been observed on cooling cycle, as illustrated by Fig. 3, which shows the temperature dependence of the total ionic conductivity for a heating/cooling cycle, before and after quenching, data being collected at a frequency of lo4 Hz. One may notice that an hysteresis phenomenon in Gonductivity takes place below the melting temperature of the sample. Furthermore, the quenched sample exhibits, on heating, a lower conductivity value than the non-quenched sample, from room temperature to 60°C. Such results can be explained by the following hypothesis: at room temperature, two phases coexist, one being crystalline PEO, and the other, more concentrated in KOH than 30wt.%, being the conducting phase. In the melting temperature range of the crystalline PEO (60-65”C), the sudden decrease of the conductivity can thus be related to an homogeni-

b

. .

. .

.

.

.

. .

.

.

I

0

5ooo

endo

‘

IcaxJ

15W

2cccM

Z/Ohms

Fig. 1. (a) Temperature dependence of the total ionic conductivity of a 6Owt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte. (b) Impedance plots of a PEO 6Owt.%, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte between two nickel electrodes, at different temperatures: (a) 24°C (A) 64”C, (0) 94°C. Frequency range: 0.1 Hz to 2 MHz.

0

20

40

60

Temperature

80

loo

(“C)

Fig. 2. DSC trace of a 6Owt.% PEO, 30wt.% KOH and 10 wt.% H,O solid polymer electrolyte at a heating rate of 3°C min- ’ (onset temperatures).

Polymer

electrolytes.

Application

2451

to nickel batteries

l.

. . i

.

4

4

2.5

2.7

3.1

23 looo/r

3.3

2.5

3.5

2.7

239

3.1

3.3

3.5

(mom/K-l

(l/K)

Fig. 4. Variation of the total ionic conductivity vs. reciprocal temperature of a 50 wt.% PEO and 50 wt.% KOH solid polymer electrolyte. -2,5 b

KOH in the composition. As far as we know, no value higher than 10-3Scn-1 at room temperature has already been obtained for alkaline PEO-based solid polymer electrolytes. Application: prospective study of the cyclability of nickel secondary batteries in the water-containing composition of 60 wt.% PEO, 30 wt.% KOH and

-5.5 2.5

2.7

2.9 Iccwr

3.1

I

I

3.3

3.5

(l/K

Fig. 3. Temperature dependence of the total ionic conductivity of a 6Owt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte (data collected at a frequency f = lo4 Hz); (a) before quenching; (b) after quenching.

zation in KOH concentration. This is consistent with experiments previously heating/cooling the described. Indeed, it is well known that the crystallinity of a polymeric sample is less after quenching. This has also been confirmed by experiments carried out with plasticized compositions showing that the decrease in conductivity over the temperature range 6t-65°C is less important and that this effect is more pronounced as the amount of plasticizer increases.

lOwt.% H,O The Nickel/Cadmium 6Owt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte secondary battery tested was manufactured by assembling vertically 11 cadmium and 12 nickel circular electrodes. Each positive electrode (thickness: 7.5 x 10e4m and area 6.16cm’) is based on nickel hydroxide Ni(OH),, and has a theoretical capacity of 250Ahcme2. Each negative electrode (thickness: 5.5 x 10e4m and area 6.16cm2) is based on cadmium electro-deposited on a nickel-plated iron and has a theoretical capacity of support, 550Ahme2. The separator is a polyamide knitted sheet (thickness: 4 x 1O-4 m). The limiting electrode

1

Anhydrous composition of 50 wt.%PEO and 50 wt.%

KOH Figure 4 shows a plot of log (a) vs. l/T for this composition. This curve also exhibits three domains. However, in the 60-65°C temperature zone, the fall of the conductivity value is much less pronounced. These results are consistent with the previously suggested hypothesis, and with the DSC analysis reported in Fig. 5, from which the crystallinity was determined; it was found to be lower than that of the 6Owt.% PEO, 30wt.% KOH and lOwt.% H,O composition. On the other hand, conductivity values of the anhydrous composition are much higher, and such good conductivity over the whole temperature range can be related to the high concentration of

endo

I 64 “C

-1 0

20

40

60

80

100

Temperature (“C) Fig. 5. DSC curve for a 50wt.% PEO and 50 wt.% KOH solid polymer

electrolyte at a heating (onset temperatures).

rate of 3”Cmin-t

2452

J. F. FAUVMQUE et al.

is the positive one, and the element has thus a theoretical capacity of 1.7Ah (based on observed data in an aqueous KOH solution). The reactions involved in the cycling process of the element are the following: Ni(OH), + OH-

1

charge and discharge curves are represented in Fig. 6a and b, respectively, showing a well-defined plateau at 1.1 V for the discharge, and at 1.35 V for the charge. Cycling experiments have been performed between 0.8 V and 1.6 V, at a current density value of 2.2Am-*. The average capacity of the element obtained up to now is 1.1 Ah, ie about 65% of the theoretical capacity previously mentioned. Thus, the actual capacity of the element is assumed to be 1.1 Ah. These prospective results indicate the rather good cyclability of both nickel and cadmium electrodes in the 60wt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte. Cycling life of this secondary battery is still under evaluation. Prototypes of higher capacity will be subsequently investigated, and the study will be extended to the

NiOOH + e- + H,O

dirsbarge charge

Cd(OH), + 2e- L

Cd+20H-

discharge

Cell testing was done using a computer controlled constant current cycler. The cycling performance for polymer electrolyte Nickel/Cadmium cell, illustrated by the faradic yield vs. number of cycles, and typical

1.8 a

1,7 1-6 >

1.5

8

194

2

1.3

;

1,2

u

1,l

1

0

0.9 68

0

10

20

30

40

Time ! h

196 a

1.5 1,4 > 1,3

0.8 0,7 66

d

’

.

10

5

’

’

15

20

’

2.5

r

’

30

35

Time I h

-i 0

5

10

15

20

25

Cycle number

Fig. 6. (a) Typical charge and discharge curves of a 60 wt.% PEO, 30 wt.% KOH and 10 wt.% H,O solid polymer electrolyte Nickel/Cadmium cell. (b) Faradic yield vs. cycle number for a 60 wt.% PEO, 30 wt.% KOH and 10 wt.% H,O solid polymer electrolyte Nickel/Cadmium cell.

Polymer electrolytes.

Application

2453

to nickel batteries

of 160 Ah rne2. The reactions involved in the cycling process of the element are the following: clWpe NiOOH + e- + H,O Ni(OH), + OH- L discharpe charge

ZnO + 2e- + H,‘O ,I

Zn + 20H-

disclurpe

Cell testing was performed using a CEAMID-6 type chrono-amperostat. This element exhibits a good discharge behaviour, even under a rate of C/8. It was tested in cycling experiments under the following conditions: charge rate = C/16, discharge depth = 81% of the capacity, discharge rate = C/8. Its internal resistance value is 6 Q. At present, one of such cells has been cycling for 60 cycles, without any short-circuit (major problem encountered in aqueous media). A decrease of capacity has been observed from the 47th cycle on. The characteristic curves for charge and discharge of the nickel and zinc electrodes prove their good cyclability, with faradic yields up to 87%. All these results are presented on Fig. 7ad

Time I h

I 10

15

20

25

Time / h

CONCLUSION 100,

1

.Q J

40. 20. OTd 10

*a .

20

30

40

50

l*,, 60

Fig. 7. (a) Typical charge and discharge curve of a 60 wt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte-Nickel electrode. (b) Typical charge and discharge curve of a 6Owt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte-Zinc electrode. (c) Typical charge and discharge curve of a 6Owt.% PEO, 30 wt.% KOH and 10 wt.% H,O solid polymer electrolyteCell voltage. (d) Variation of the capacity vs. cycle number for a 60 wt.% PEO, 30 wt.% KOH and 10 wt.% H,O solid polymer electrolyte.

cyclability

of other

nickel

secondary

batteries,

such

as Nickel/Metal hydride. A Nickel/Zinc 60wt.% PEO, 30wt.% KOH and lOwt.% H,O solid polymer electrolyte secondary battery has been elaborated and tested in such medium. It has been manufactured by assembling a central circular zinc electrode, between two nickel electrodes. The negative electrode (thickness: 2 x 10W4m and area 5 x 10m4m2) is an expanded zinc grid, welded on each side on a cadmium drilled sheet as collector, and has a theoretical capacity of 820Ahm-‘. The positive electrodes (thickness: 10m4m and area 5 x 10-4m2) are made of nickel hydroxide Ni(OH), , and have a theoretical capacity

It was presented a new class of solid polymer electrolytes, based on PEO and KOH, exhibiting an unusual conductivity vs. temperature behaviour which can be attributed to the existence of two phases within the electrolyte, one of these being the conducting, high concentrated phase. At the same time, both compositions have significant ionic conductivity values, at room temperature as well as at high temperature. The water-containing one can easily be used in nickel secondary batteries, regarding the good cyclability of Nickel/Zinc and Nickel/ Cadmium cells. Acknowledgements--The authors thank the Agence de I’Environnement et de la Ma&rise de I’Energie (Ademe) and ElectricitC de France (EDF) for financial support.

REFERENCES 1. D. E. Fenton, J. M. Parker and P. V. Wright, Polymer 14, 589 (1973). 2. M. B. Armand, J. M. Chabagno and M. J. Duclot, Second International Conference on Solid Electrolytes, St Andrews, paper 6.5 (1978). 3. M. B. Armand, J. M. Chabagno and M. J. Duclot, in Fast Transport in Solids (Edited by P. Vashishta), p. 131-136, North-Holland, New-York (1979). 4. Polymer Electrolytes Reviews, Vols 1 and 2 (Edited by J. R. MacCallum and C. A. Vincent). Elsevier Applied Science, London (1987) and (1989). 5. M. Armand, W. Gorecki and R. Andreani, Second Int. Symp. on Polymer Electrolytes (Edited by B. Scrosati), p. 91-97, Elsevier, Amsterdam (1990).