Development of a rechargeable zinc-air battery

Transcription

1 1.1149/ The Electrochemical Society Development of a rechargeable zinc-air battery Gwenaëlle Toussaint* a, Philippe Stevens a, Laurent Akrour a, Robert Rouget b and Fabrice Fourgeot b a EDF R&D, Avenue des Renardières, Morêt-Sur-Loing Cedex, France b SCPS,85-91 Boulevard Alsace Lorraine, Rosny sous bois Cedex, France An electrically rechargeable zinc-air cell was developed and demonstrated using a bi-electrode on the cathode side and a 3D zinc electrode. About 2 cycles corresponding to 5h of operation was achieved with this configuration. An innovating hybrid bi-electrode was evaluated which significantly increase the energy efficiency of our system to about 7% with an energy density close to 11 wh/kg and also improves the power response of the zinc air battery for punctual power demand application. Introduction The limitation to the commercial development of the electric vehicles (EV) can be attributed for most part to the battery. The cost of lithium ion batteries is too high and the energy density of cheaper batteries such as the lead-acid battery is too low. These batteries also suffer from environmental and safety problems. Zinc air batteries have very high theoretical energy densities (> 9 Wh/kg) and are made of cheap abundant materials (zinc, potassium hydroxide, carbon, manganese). The cost of zinc is close to 1 /kwh based on current zinc prices (around 1 /tonne) (1) and the air electrode can be made for as little as 18$/m² for the equivalent of 5 vehicles/year (2), which adds around 4-5 /kwh to the cost of the battery. Zinc-air batteries could potentially be manufactured at a cost lower than lead-acid batteries, the cheapest technology today. World resources of zinc are estimated at 3 times greater than those of lithium and therefore pose no threat of resource limitation, even if all vehicles were powered with zinc. The zinc-air battery is an aqueous system with no toxic or inflammable substances. Zinc is also easily recyclable. They are therefore safe and environmentally friendly. Commercial primary button cell zinc-air batteries are sold by the millions everyday for hearing-aids. They can have very high energy densities, up to 442 Wh/kg for the Duracell DA675 button cell. The primary zinc-air battery is therefore a mature technology, which is not the case for the secondary, rechargeable variety. All these properties therefore make them very good candidates for electric vehicles. Unfortunately, their major drawback is the very low lifetime of the battery when recharged electrically. The rechargeable zinc-air battery suffers from two important weaknesses which does not affect the primary battery: the poor stability of the air electrode when used to charge the battery and the formation of dendrites on the zinc electrode leading to short circuits and shedding of zinc. Downloaded on to IP address. Redistribution subject to ECS 25 terms of use (see ecsdl.org/site/terms_use)

2 Positive electrode O 2 + H 2 O + 4e - 4OH - E =,41 V/ENH [1] The causes of failure of the positive electrode are multiple: Mechanical failure: During discharge, the zinc-air cell consumes oxygen from the air at the positive electrode. An electrode with a very high effective surface area is therefore preferable for the gas to react. The electrochemical reaction can only occur if, the reactant gas (oxygen) is in contact with the liquid electrolyte over this high surface area, but also with an electronic conductor for current collector. The air electrode is therefore designed to work with a triple phase interface composed of a gas, a liquid and a solid (3). To reduce electrochemical losses, a catalyst also needs to be present at each triple phase reaction points. These triple phase reaction interfaces are obtained by using a porous, 3D electrode into which both air and the liquid electrolyte can penetrate. The porous structure is composed of small carbon particles onto which a catalyst is deposited, held together by a PTFE polymer. This mixture is pressed onto a metal grid, which acts as a current collector. Concentrated aqueous potassium hydroxide is used as an electrolyte. This fragile structure optimised for gas-liquid and gas-solid interfaces gives very good results, with current densities as high a 1 A/cm² when used in fuel cells (3). When used in reverse to charge the battery, oxygen is not consumed but evolved from the electrode. The reaction is now between a liquid (the electrolyte) and a solid (the electronic conductor) to produce a gas (oxygen). It occurs at two phase interface which does not require very high surface areas since the surface concentration of reactant is much higher than in a gas. The porous structure of the air electrode is not optimised for gas evolution. The mechanical pressures, which the carbon structure is subjected to as a result of the gas, produced trying to escape from the electrode, causes mechanical breakdown of the electrode (Fig 1b). The electrode then rapidly drops in performance when used in discharge mode as an air electrode again. e - e - O 2 OH - OH - O 2 O 2 O 2 Zn (a) (b) Figure 1: Air electrode during discharge (a) and charge (b) Zn Electrochemical failure: The open circuit voltage of the zinc-air cell is typically around 1.5V and drops to 1.2V when in operation in discharge. These potentials are Downloaded on to IP address. Redistribution subject to ECS 26 terms of use (see ecsdl.org/site/terms_use)

3 compatible with the air electrodes and do not induce pronounced corrosion or breakdown of the electrode materials. However, the cell voltage needs to be raised to around 2V during the charging process due to the large overpotential of the oxygen evolution reaction. These very high oxidation potentials have two consequences: (a) MnO 2 which is used as a catalyst is not stable in alkaline solution under these potentials and is oxidised to MnO 4 - which goes into solution (4). (b) Carbon corrosion by oxidation (5) is accelerated under these potentials Reaction with carbon dioxide: Carbon dioxide from the air will react with the alkaline electrolyte (KOH) to produce potassium carbonate which will eventually precipitate. This has two negative effects: carbonate precipitation inside the air electrode blocks the pores of the electrode leading to irreversible failure of the electrode and the reaction leads to reduction of active material by consuming the electrolyte. Negative electrode Zn + 4 OH - Zn(OH) e - Zn(OH) -2 4 ZnO + 2 OH - + H 2 O E = -1,25 V/ENH [2] During cycling, zinc will not necessarily be deposited where it has been consumed. As a result, the zinc electrode will start to change shape to produce an electrode with uneven thickness. The restructuring of the electrode will lead to the production of zones in the zinc electrode, which become less active, resulting in loss of performance. This process ultimately produces shedding of active material and irreversible loss of capacity. Another type of shedding originates from the growth of zinc dendrites which can break-off, or worse, grow to form a short-circuit with the positive electrode. Zinc air battery for electric vehicles (EV) The EV application requires high energy density batteries in order to obtain long driving ranges on a single charge. Power density is also important in order to meet the vehicle requirements when accelerating and going uphill, during shorter periods of time. Metal-air batteries are known to have poor power densities, mainly due to the performance of the air electrode. Metal-air batteries also have poor energy efficiencies due to the nature of the oxygen reduction/evolution reaction. Both the ORR and OER have high overpotentials. The poor efficiency can have an important impact on the energy saved during regenerative breaking. In this paper solutions to prevent degradation of the air electrode, increase its power and its energy efficiency are described. Improving zinc-air battery cycling lifetime Reversibility of the air electrode : positive bi-electrode A secondary oxygen evolution electrode was used which effectively decouples the air electrode when the zinc-air battery is charged. This has the advantage that standard air electrodes using cheaper catalysts such as MnO 2 can be used. The air electrode is used for what it was designed for and optimised, and its lifetime is not affected. A light oxygen Downloaded on to IP address. Redistribution subject to ECS 27 terms of use (see ecsdl.org/site/terms_use)

4 evolution electrode was used made of a metallic grid, which does not add any significant weight to the battery. Reversibility of the Zinc electrode : 3D electrode A zinc electrode developed by SCPS was used which is designed to reduce the formation of dendrites in Ni-Zn battery configuration (6) and which enables 8 to 15 cycles to be obtained without degradation. It is composed of a metal foam current collector filled with a mixture of zinc, ceramic electronic conductor (TiN) and a polymer binder. The ceramic conductor also helps to retain zincates ions, allowing a homogeneous zinc deposition in charge. Experimental The bi-electrode concept on the cathodic side was evaluated in a half cell configuration containing the combination of an air electrode used in discharge and the evolution electrode used in charge. A Hg/HgO-KOH (1M) reference electrode was used and a counter electrode made of a stainless steel metallic grid. The tests were performed in a concentrated potassium hydroxide (7 M). Commercial air electrodes provided by Electric Fuel Ltd and elaborated with catalysts based on Cobalt oxide (E5 type) or Manganese dioxide (E4 type) were used. The evolution electrodes tested are essentially made of a stainless steel 316L type grid (5 mesh). For the study on the electrochemical stability of the air electrode, the air electrodes were subjected to different oxidation potentials in a half cell configurations using a Hg/HgO- KOH (1M) reference electrode and a platinum counter-electrode in 7M KOH solution. The different oxidation potentials were applied for 2 minutes after which a sample of electrolyte was taken for UV-Vis spectroscopy measurements. The samples of electrolyte were returned to the half cell after each measurement. Air electrodes were made with Vulcan XC72 carbon with and without MnO 2 catalyst for its experiment. Complete zinc air cell tests were tested in a symmetric configuration using a zinc electrode in a central position surrounded by two bi-electrodes (Figure 2). A specific proprietary electrolyte formulation developed by SCPS was used essentially composed of concentrated potassium hydroxide (7 M). Zinc electrodes with dimensions 5 cm x 6 cm were manufactured with different capacities (3, 6 and 9 Ah). To prevent carbonate formation, the air fed to the air electrodes was first bubbled through a KOH solution. A relay was used to connect the appropriate cathode in charge or in discharge. The electrochemical tests were performed with a multichannel VMP3 potentiostat made by Biologic. Downloaded on to IP address. Redistribution subject to ECS 28 terms of use (see ecsdl.org/site/terms_use)

5 () (-) (+) air air air electrode zinc electrode oxygen evolution electrode KOH ( 7M) air air Figure 2: Schematic of Zinc air cell Results and Discussion The air electrode was cycled in both oxygen reduction and oxygen evolution (Figure 3).,8 E air ( V vs. Hg/HgO KOH 1M),6,4,2 -,2 -,4 -,6 charge on air electrode discharge on air electrode -, Cycles Figure 3: Cycling of air electrode (2 min +/- 1 ma/cm² cycles) The performance of the electrode can be seen to degrade very rapidly, after only a few cycles. After ten cycles, the electrode has almost no activity the remaining performance can be attributed to the catalytic activity of the carbon. The degradation of MnO 2 catalyst was demonstrated using UV-Vis Spectroscopy. The UV-Vis Spectra of the electrode without MnO 2 catalyst shows no change with potential, except for a small peak around 37 nm (Figure 4). The same experiment using an electrode made with XC72 carbon containing MnO 2 catalyst clearly shows two absorption peaks at 44 and 61 nm (Figure 5) which can be attributed to the degradation and dissolution of the MnO 2 catalyst at potentials equal or greater than.5v vs Hg/HgO/KOH(1M). Downloaded on to IP address. Redistribution subject to ECS 29 terms of use (see ecsdl.org/site/terms_use)

6 1 T (%) wave length (nm),4 V vs Hg/HgO,45 V vs Hg/HgO,5 V vs Hg/HgO,55 V vs Hg/HgO,6 V vs Hg/HgO,65 V vs Hg/HgO,7 V vs Hg/HgO,75 V vs Hg/HgO Figure 4: UV-Vis spectra of an air electrode without catalyst in 7M KOH at different potentials 1 8 T (%) wave length (nm),4 V vs Hg/HgO,45 V vs Hg/HgO,5 V vs Hg/HgO,55 V vs Hg/HgO,6 V vs Hg/HgO,65 V vs Hg/HgO,7 V vs Hg/HgO,75 V vs Hg/HgO Figure 5 : UV-Vis spectra of an air electrode with MnO 2 catalyst in 7 M KOH at different potentials. By using the bi-electrode system, which uses a secondary oxygen evolution electrode, the air electrode reached up to 3 cycles. Very little degradation is observed, even after 5 cycles. Downloaded on to IP address. Redistribution subject to ECS 3 terms of use (see ecsdl.org/site/terms_use)

7 E + (V / Hg/HgO KOH 1M),8,6,4,2 -,2 -,4 -,6 charge on stainless steel electrode O discharge on air electrode -, cycles Figure 6 : Cycling of bi-electrode (1 ma/cm², 2 hour cycles) The half cell test showed that the bi-electrode system was very effective in preventing air electrode degradation. The bi-electrode was then tested in full cell configuration with a zinc electrode in the symmetrical cell shown in Figure 2. Cell voltage (V) -,6 -,7 -,8 -,9-1 -1,1-1,2-1,3-1,4-1,5-1,6-1,7-1,8-1,9-2 -2,1-2,2 5 time /h Figure 7 : Charge/discharge cycle on a 9 Ah cell 1 The 9 Ah cells (Figure 7) can be cycled with a faradaic efficiency of between 1 and 9%. Low current densities were used, between 5 and 15 ma/cm², and energy efficiency is just above 5%. An energy density of 84 Wh/kg (without the casing) was achieved with 6Ah cells and with an excess of electrolyte. By correcting for the excess electrolyte, more than 1 Wh/kg can be obtained and 13 Wh/kg if using a 9 Ah zinc electrode. With the best cells, more than 2 cycles were obtained and a lifetime of 5 hours. This is still low for the EV application but these are encouraging results compared to the state of the art for an electrically rechargeable zinc-air battery. Progressive loss of capacity and faradaic efficiency was also observed which could be attributed to the shedding of zinc and the formation of dendrites. The degradation of the zinc electrode was accelerated when using high current densities and important overcharge. Preliminary studies have shown that the nature of the oxygen evolution electrode is partially Downloaded on to IP address. Redistribution subject to ECS 31 terms of use (see ecsdl.org/site/terms_use)

8 responsible for this degradation, and an improved evolution electrode is under development. Improving power density and energy efficiency The oxygen evolution electrode is responsible for most of the energy loss in cycling. The addition of an oxygen evolution catalyst such as Ni, Co or Ni/Co mixed oxides (7) reduces the oxygen evolution overpotential but only to a limited extent. A different approach was experimented in which the secondary electrode is not only used as an oxygen evolution electrode but also as capacitive faradaic electrode, which can be used in both charge and discharge. This electrode can also provide extra power for a short duration of time. In this hydride zinc-air cell, during discharge, the capacity of the secondary electrode is first discharged at a higher potential, thus providing high power and energy efficiency. Once it drops to the air electrode potential, the air electrode is used to enable the full capacity of the zinc-air cell to be delivered. During charge the capacity of the secondary electrode is first restored at a lower potential and oxygen evolution on this electrode then enables the zinc electrode to be fully charged at a higher potential. Short charge discharge cycles can also be obtained without using the air electrode or using the electrode in oxygen evolution at much higher energy efficiencies. This feature is particularly interesting for regenerative breaking. Experimental In order to increase the power density and energy efficiency of the zinc-air battery, a hybrid cell was developed in which the oxygen evolution electrode is replaced by a nickel electrode of the type used in Ni-Zn batteries. The Ni/Air electrode association was tested in a half cell test configuration as described above where the stainless steel grid was replaced by a capacitive nickel electrode. This electrode remains connected during charge and discharge. The air electrode is connected to the second electrode during the discharge step. The nickel electrode with dimension 5 cm x 6 cm was used. The electrode essentially consists of nickel oxide as NiO-NiOOH. Holes of 2 mm diameter were made to enable the electrolyte continuity through this compact electrode. A complete zinc air symmetric cell using the modified bi-electrode integrating the hybridization concept was tested with a 3 Ah zinc electrode. Results and Discussion The cycling curve obtained with a hybrid bi-electrode in a half cell configuration presents a characteristic discharge curve with two stages (Figure 8). Downloaded on to IP address. Redistribution subject to ECS 32 terms of use (see ecsdl.org/site/terms_use)

9 ,5 charge Ewe ( V vs. Hg/HgO 1M),4,3,2,1 -,1 -,2 -, time/h discharge Figure 8: cycling of the hybrid bi-electrode (6 ma/cm², 12 hours cycle) At the begin of the discharge, the nickel oxide reduction takes place at high potential corresponding for the first stage ( ). In the second phase, when the nickel electrode is completely discharged, a second stage is observed ( ) with a leveling of the potential that of the air electrode. With this hybrid bi-electrode, the initial polarization in discharge is reduced which will improve the energy efficiency and a power density of the zinc air cell. 2 Cell voltage (V) -,8-1 -1,2-1,4-1,6 discharge OCV -1,8-2 -2,2 charge 1 15 time/h 2 Figure 9: Charge/discharge cycle on a 3 Ah cell with a hybrid bi-electrode (3 ma/cm²) The hybridization concept was tested in a complete zinc air cell configuration. The discharge curve observed on the first cycles are similar to that observed in the half cell tests with the two characteristic phases (Figure 9). The duration of the first stage is long due to the capacity of the nickel electrode that was chosen. Here, the nickel electrode was not optimized in term of capacity and was oversized to demonstrate the advantages of this concept. Downloaded on to IP address. Redistribution subject to ECS 33 terms of use (see ecsdl.org/site/terms_use)

10 The 3 Ah cell was cycled with a faradaic efficiency of 96% and a higher current density was applied (3 ma/cm²). The electrical performance obtained here are interesting. The average voltage in charge is about 2 Volt and close to 1,8 Volt in discharge instead of 1,2 Volt which increases the cycle energy efficiency to 7%. If using a 9Ah zinc electrode, an energy density of 11 Wh/kg (without the casing) would have been achieved. A greater energy density is expected with the use of a nickel capacitive electrode correctly sized. Over 9 cycles were obtained with this cell. Slow degradation of the air electrode was observed during the cycling test attributed to the high oxidation voltage of the air electrode induced by the nickel electrode connected at the begin of the discharge. An optimized management of the hybrid bi-electrode should avoid this failure. Conclusion The problems of air electrode degradation when used as a reversible electrode were solved in this study by using a bi-electrode in which the charge and discharge functions are separated and an electrically rechargeable zinc-air cell was successfully demonstrated. More than 5 h of operation and 2 cycles were achieved with our best cell. Some improvements are still needed in terms of current density, energy density and cycling to be fully compatible with an EV application. Nevertheless the target objectives of a battery for EV application could be revised in light of these results. Indeed the development of a potentially low cost zinc air battery could enable a replacement of the battery pack during the lifetime of the car. The lifetime of the battery could therefore be shorter than that of the car. With this approach the performances obtained here are very encouraging and should already give a lifetime for the battery of several years for a high capacity battery (eg 1-2 charges per week). In addition, the hybridization concept presented in this study gives us an interesting and simple alternative to improve power and energy efficiency of the zinc air technology by using a faradaic capacitive second electrode. The preliminary tests of the hybrid bielectrode show an increase of the energy efficiency of 7 % and an energy density greater than 11 Wh/kg is expected. Applications requiring high power for a short duration of time such as regenerative breaking and acceleration can be now considered with this system. References E.J. Carlson, P. Kopf, J. Sinha, S. Sriramulu, and Y. Yang, Cost Analysis of PEM Fuel Cell Systems for Transportation, Tiax llc report NREL/SR , U.S. Department of Commerce National Technical Information Service (25) 3 S. Srinivasan, R. Mosdale, Ph. Stevens, C. Yang, Annu. Rev. Energy Environ, 24, 281 (1999) 4 M. Pourbaix, Atlas of electrochemical equilibria in aqueous solutions. 2d English, Houston, Tex.: National Association of Corrosion Engineers (1974). 5 L. M. Roen, C. H. Paik, T. D. Jarvi, Electrochem. Solid-State Lett, 7(1), A19-A22 (24) 6 B. Bugne, D. Doniat, R. Rouget, Patent WO 24/1364 A2. 7 C. Boccaa, A. Barbuccia, M. Delucchia and G. Cerisola, Int. J.of Hydrogen Energy, 24(1), 21 (1999) Downloaded on to IP address. Redistribution subject to ECS 34 terms of use (see ecsdl.org/site/terms_use)