Scholarly article on topic 'Rechargeable lithium–air batteries: characteristics and prospects'

Rechargeable lithium–air batteries: characteristics and prospects Academic research paper on "Materials engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Materials Today
OECD Field of science
Keywords
{}

Abstract of research paper on Materials engineering, author of scientific article — Nobuyuki Imanishi, Osamu Yamamoto

High specific energy density batteries are attracting growing attention as possible power sources for electric vehicles (EVs). Lithium–air batteries are the most promising system, because of their far higher theoretical specific energy density than conventional batteries. However, no technical basis exists to support the high energy density estimated from calculation. In this review, we will discuss the state-of-the art of lithium–air (or oxygen) batteries, as well as prospects for the future, with a focus on materials.

Academic research paper on topic "Rechargeable lithium–air batteries: characteristics and prospects"

ELSEVIER

30 O X

Rechargeable lithium-air batteries: characteristics and prospects

Nobuyuki Imanishi and Osamu Yamamoto*

Graduated School of Engineering, Mie University, 1577 Kurimamachiya-cho, Tsu, Mie 514-8507, Japan

High specific energy density batteries are attracting growing attention as possible power sources for electric vehicles (EVs). Lithium-air batteries are the most promising system, because of their far higher theoretical specific energy density than conventional batteries. However, no technical basis exists to support the high energy density estimated from calculation. In this review, we will discuss the state-of-the art of lithium-air (or oxygen) batteries, as well as prospects for the future, with a focus on materials.

Introduction

Recently, many research groups have been developing rechargeable lithium-air batteries, because of the far higher theoretical energy density than that of conventional batteries. These high specific energy density batteries have the potential for application as power sources in electric vehicles (EVs). At present, two types of reversible lithium-air batteries have been proposed;namely nonaqueous [1] and aqueous systems [2]. The specific energy density of the non-aqueous system calculated from the reaction:

2Li + O2 = U2O2 (1)

and the open circuit voltage (OCV) of 2.96 V [3], are 3460 Wh kg"1 and 6940 Wh L"1 for the discharge state. For the charged state, oxygen is excluded and the specific energy density is 11,680 Wh kg"1, which is close to the energy density of gasoline (ca. 13,000 Wh kg"1). The energy density per unit mass and per unit volume of non-aqueous lithium-air batteries is about 10 times and 6 times higher than those of lithium-ion batteries, respectively, with a carbon anode and LiCoO2 cathode. In an aqueous system, water molecules are involved in the redox reaction at the air cathode:

4Li + 6H2O + O2 = 4(LiOH • H2O) (2)

The cell voltage is dependent on the concentration of OH" in the solution and is 3.90 V in a neutral solution [4]. However, the concentration of OH" increases with discharge. Saturation of LiOH is reached at 5.3 ML"1, which corresponds to a discharge depth of (on ca.) 5%. The reaction product is deposited in the

*Corresponding author: Yamamoto, O. (Yamamoto@chem.mie-u.ac.jp)

electrolyte as LiOHH2O. The OCV with a LiCl and LiOH saturated aqueous solution was around 3.0 V [5], where LiCl is saturated to prevent the reaction of LiOH and the lithium protective layer of a water stable lithium ion conductor as discussed later. The specific energy density of the aqueous system calculated from reaction (2) and OCV of 3.0 V is 1910 Wh kg"1 for the charged state and 2004 Wh kg"1 for the discharge state. The mass specific energy density is five times higher than that of a conventional lithium-ion battery.

Nissan Motors, Japan, has commercialized EVs that employ lithium-ion batteries (the energy density is about 100 Wh kg"1), of which the driving range was announced to be 200 km with a full charge, although the driving range may be reduced when the air conditioning system is used [6]. The battery capacity is 24 kWh and the weight of the batteries is approximately 15% of the total car weight. The driving range is too short compared to that of internal combustion engine vehicles. The energy density of a battery that is comparable to that of an internal conversion engine would be approximately 700Whkg"1 [7]. The practical energy density of batteries is related to the calculated energy density by multiplying by a conversion factor;for example, those of a large size lithium-ion battery and a nickel metal hydride battery are 0.26 and 0.32, respectively, which were calculated from the theoretical energy density estimated from the capacity of active materials and the reported battery energy density. Therefore, to obtain practical batteries with an energy density of 700 Wh kg"1, the calculated energy density should be more than 2000 Wh kg"1. The nonaqueous lithium-air system is an attractive candidate for batteries in EV, although aqueous lithium-air batteries could also be candidates from the point of view of the energy density.

1369-7021/06/$ - see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/imattod.2013.12.004

The first report of rechargeable lithium-air cells was in 1996 by Abraham and Jiang [8]. They used a lithium metal anode, a lithium conducting gel-type polymer electrolyte, and a carbon composite electrode with a catalyst. The cell could be recharged with a good columbic efficiency. Over the following ten years, research activities into rechargeable lithium-air batteries were limited. However, since Bruce and co-workers revisited lithium-air cells composed of a lithium metal anode, a non-aqueous electrolyte of 1 M LiPF6 in propylene carbonate (PC), and a porous cathode with manganese dioxide in 2006 [9], research activity on lithium-air batteries has grown. However, no technological basis exists to support the highly optimistic energy densities projected for lithium-air batteries. Moreover, capabilities for high power density and extended deep cycling have not been shown. In this review, an up to date perspective on lithium-air battery prospects for EV applications will be introduced, with an emphasis on the materials for lithiumair batteries.

Electrolytes for non-aqueous systems

The requirements for the electrolyte in the non-aqueous lithiumair system are as follows: (1) stable with lithium metal, (2) a high oxidation potential, (3) a low vapor pressure and high boiling point, (4) a high lithium salt solubility and a good chemical stability. Abraham and Jiang used a gel-type polymer electrolyte of LiPF4 in polyacrylonitrile (PAN) with ethylene carbonate and PC [8]. These types of electrolyte have been used for gel-type lithiumion batteries [10]. The cell was recharged with good coulombic efficiency for two cycles. Following this work, Read et al. studied the discharge capacity and rate capacity of various non-aqueous electrolytes, including carbonate and ether based electrolytes, and found that electrolyte formulation has the large effect on capacities [11-13]. They found that the solubility of oxygen in the electrolyte and the diffusion coefficient calculated from the viscosity have a direct impact on the discharge capacity and rate capacity. The cyclic performance of a 1-MnO2/Super P carbon black air cell in a 1 M LiPF6-PC-dimethyle ether electrolyte between 4.15 and 2.0 V at 0.1 mAcm-2 was examined [11]. The cyclic performance was poor;the first cycle discharge capacity of 85 mAh g-carbon (g-c)-1 failed at 25 mAh g-c-1 after three cycles. In 2006, Bruce and coworkers [9] reported more excellent cyclic performances for the cell with a similar electrolyte of 1 M LiPF6 in PC and Super P carbon black with electrolytic manganese dioxide. Fig. 1 shows the cell potential change with capacity at a rate of 50 mA g-c-1 and cyclic performance in 1 atm, O2 at rates of 75 and 100 mA g-c-1. However, a large potential difference during the discharging and charging processes was observed. A similar cyclic performance was reported for the Li/1 M Li(CF3SO2)2N (LiTFSI) in PC/C, O2 cell by Mizuno et al. [14]. The initial capacity of 800 mAh g-c-1 failed for 600 mAh g-c-1 after 100 cycles at 0.02 mA cm-2. They observed that the main reaction product was not Li2O2, but was carbonate species from the decomposition of PC. More recently, Bruce and co-workers reported that a lithium-air battery containing an alkyl carbonate electrolyte discharges by formation of C2H6(OCOLi)2, Li2CO3 HCO2Li, CH3CO2Li, CO2, and H2O at the cathode, due to electrolyte decomposition [15]. Charging involves oxidation of these compounds. The different pathways for charge and discharge were considered with the observed voltage gap shown in Fig. 1. They concluded that organic carbonates are not suitable as

FIGURE 1

(a) Variation of potential on discharge then charge corresponding to the third cycle of the cell at a rate of 50 mAg-1. Capacities are expressed per gram of carbon in the electrode. (b) Variation of discharge capacity with cycle number. Capacities are per gram of carbon. Figure from Ref. [9].

electrolytes and it is important to investigate other solvents to find a suitable electrolyte.

Luntz and co-workers [16-19] have extensively examined the stability of numerous solvent/salt combinations using differential electrochemical mass spectrometry (DEMS). The most important conclusion was that although some electrolytes have the absolute moles of O2 consumed during discharge of (e/O2) « 2.00 (e.g. ethers and CH3CN), all electrolytes have the absolute moles of

O2 evolved during charge of an (e/O2) > 2.0 and the chemical rechargeability (characterized by the oxygen recovery efficiency) is <0.9. This naturally severely limits cyclability.

In this regard, significant attention is now focused on electrolytes based on ethers [20-23]. Aurbach et al. [24] have used end-capped glymes such as tetraethylene glycol dimethyl ether (TEGDME) as the electrolyte for lithium-ion batteries. However, the stability of ethers in lithium-air cells is questionable. Bruce and co-workers concluded that electrolytes based on linear and cyclic ethers all exhibit electrolyte decomposition when used in lithium-O2 cells by combining electrochemical measurement with powder

FIGURE 2

(a) Powder XRD patterns of the composite cathode (Super P/Kynar) cycles in 1 M LiPF6 in TEGDME under 1 atm O2 between 2 and 4.6 V vs. Li/Li+. Rate = 70 mAg"1. (b) Load curves for the same cell. Figure from Ref. [21].

X-ray diffraction, FTIR and NMR spectroscopy [21]. Fig. 2 shows the XRD patterns and the charge and discharge profiles of the Li/ LiPF6-TEDME/Super P carbon black, O2 cell. More recently, Scrosati and co-workers reported a different conclusion [23]. The Li/ LiCF3SO3-TEGDME/Super P carbon black, O2 cell showed no discharge capacity degradation after 100 cycles at the high specific discharge and charge rate of 500 mA g-c"1 as shown in Fig. 3. The materials of these cells was almost same, except for the lithium salts. The reason for the different conclusions of Scrosati etal. and the other papers is not certain, the stability of TEGDME with various lithium salts should be studied using other methods, such as DEMS. The other candidate for use as the electrolyte for lithiumair cells is dimethyl sulfide (DMSO) [13,25], and recently, Bruce and co-workers reported an excellent cyclability for the Li/0.1 M LiClO3-DMSO/porous gold, O2 cell;the cell showed 95% capacity retention from 1 cycles to 100 [26].

Carbonates decompose during the operation of lithium-oxygen batteries and the decomposition of ethers is questionable, which compelled us to search for alternative electrolyte systems for lithium-air batteries. Some ionic liquids have satisfied the

FIGURE 3

Cycling response of the Li/TEGDME-LiCF3SO3/O2 battery. Rate = 500 mAg Figure from Ref. [23].

requirements for the electrolyte in lithium-air batteries, in particular the extended anodic voltage window [27-29]. Kuboki et al. used 1-alkyl-3-methylimidazolium bis(trifluoromethyl(sulfonyl) imide in primary lithium-air cells. High discharge capacities greater than 5000 mAh g-c"1 with a very low current density of 0.01 mAcm"1 have been reported [27]. The cell was operated in open air for a long period of time, which maybe due to the stability of the interface between lithium metal and the ionic liquid to moisture corrosion. More recently, Nakamoto et al. [29] have tested many types of ionic liquids as electrolytes in rechargeable lithium-O2 cells and found that a N,N-diethyl-N-methyl-N- (2-methoxyethy-l)ammonium(bis(trifluoromethanesulfonyl)imide system had the highest LiOx generation activity among the ionic liquids examined. The cell showed a high capacity of 3000 mAh g-c"1 at 50 mA cm"2, but the coulombic efficiency for the first cycle was 66%. A lithiumair cell using a solid lithium conducting electrolyte was demonstrated by Kumer et al. [30,31] which used a NASICON-type high lithium conducting solid electrolyte membrane of Li1+xAlx-Ge2"X(PO4)3. This electrolyte has a high lithium ion conductivity of 5 x 10"3 S cm"1 at room temperature, a high stability to moisture, a wide electrochemical window, and excellent thermal stability [32]. The poor rate performance resulting from the low lithium ion conductivity needs further improvement. In Table 1, the lithium-oxygen cell performances of the typical carbonate and ether-based electrolytes are summarized. The cyclic performance is poor, except of Ref. [23], where the carbon electrode packing density was as low as 0.125 g cm"3.

Air cathodes for non-aqueous systems

The performance of rechargeable lithium-air batteries with non-aqueous electrolytes is limited by the oxygen evolution and oxygen reduction kinetics, especially oxygen evolution. Oxygen is reduced by lithium ions to form Li2O2 and/or Li2O:

2Li+ + 2e" + O2 = 4Li + + 4e" + O2

Li2O2 E° = Li2O E°

2.96 V

TABLE 1

Non-aqueous electrolytes for lithium-oxygen batteries.

Electrolyte Oxygen electrode Discharge and charge rate Difference of charge and discharge voltage (V) Capacity (mAh g-c 1st cycle After cycle (cycle) Ref.

LiPF6-PC Super P carbon + aMnO2 70 mAg-c-1 1.3 1000 600 (50) [9]

LiPF6-PC-EC-PAN Chevlon carbon + Co-phthalocyanine 0.1 mAcm-2 (discharge) 0.05 mA cm-2 (charge) 0.92 520 600 (2) [8]

LiTFSI-PC Super P carbon + MnO2 0.02 mA cm-2 1.8 800 600 (100) [14]

LiPF6-TEGDME BP2000 carbon black 0.13 mAcm-2 1.4 800 200 (5) [20]

LiPF6-TEGDME Super P carbon 70 mAg-c-1 1.8 3000 1500 (5) [21]

LiCF3SO3-TEGDME Carbon on carbon paper 500 mA g-c-1 1.7 3000 3000 (50) [23]

The analysis of the reaction products by Raman spectroscopy and XRD has revealed Li2O2 as the major discharge product [8,9], while oxygen consumption measurements during the discharge suggest a partial formation of Li2O [18]. Many studies [1] suggest that the reaction (3) is not direct, but divided into three steps as follows:

O2 + e-= O2~ O2~ + Li+ = LiO2 2LiO2 = Li2O2 + O2

(6) (7)

These different pathways for the oxygen reduction and evolution reactions result in the high voltage gap for the charge-discharge potentials, as shown in Fig. 1. The formation of peroxide ions is the observed decomposition of the solvents in the non-aqueous electrolytes. The discharge profile of a lithium-oxygen cell exhibits a plateau near 2.6 V while the charge plateau occurs at a higher potential, which is dependent on the catalyst used. Fig. 4 shows the cell potential change with charging capacity at 25 °C and 70 mAg-c-1 [33]. A strong correlation between the rate of H2O2 decomposition and the charging voltage of lithium-air cells has been observed;the highest H2O2 decomposition rate coming with the lowest charging voltage, as low as 3.15 V vs. Li+/Li for high

-1-1-1-----1 ---J-«—7-1-

no catalyst / ..-f :

[ \ .............-Z&2&JJ

J.****......"••"rH.VMimti"»«»!

L-'-v...................... N10............/

ft .................. ...........•'**" / i

{: ....... ..........

1 • ........ 'àop\............................

|i EMD..... .................. f........

UA-.V.-.^'*"'"'' ' " " ZMnÖ'b"" ...........................

it.,,,,,,.............m,,«,,....»' • ^MnQ ^

AE° = 3.1V vs. Ll*/U"

Capacity (mAh/g of Li O )

FIGURE 4

First galvanostatic change (i.e. Li2O2 evolution) for various Li-O2 cells. T = 25 °C, i =70 mAg-1 in 1 M LiPF6 in PC. Figure from Ref. [33].

surface area. Shao-Horn and co-workers have studied the catalytic activity of platinum and gold for oxygen reduction and evolution reactions in LiClO4 in PC-DME and found Au/C (Vulcan) had the highest discharge activity, while Pt/C exhibited extraordinarily high charging activity [34]. They also reported that Pt/Au/C exhibited bifunctional catalytic activity for oxygen reduction and evolution reaction kinetic in lithium-oxygen cells [35].

Carbon has been widely used as the base of porous cathodes for lithium-air cells. However, Bruce and co-workers have reported carbon decomposition. They have measured CO2 evolution from Li/LiClO4-DMSO/isotropically labeled carbon/O2 and Li/LiPF6-TEGDME/isotropically labeled carbon/O2 [36]. Fig. 5 shows the content of 12CO2 and 13CO2 at different states of charge and discharge, where 12CO2 comes from the decomposition of the electrolyte and 13CO2 from the decomposition of carbon. These results suggest that the electrolytes of LiClO4 in DMSO and LiPF6 in TEGDME and the carbon electrode decompose. The carbon electrode decomposes above 3.5 V to form Li2CO3. They also claimed that hydrophobic carbon is more stable and less able to promote electrolyte decomposition than its hydrophilic counterpart. The DMSO based electrolyte is somewhat more stable than TEGDME. The carbon decomposition by the oxygen evolution reaction was also observed for the aqueous electrolyte. Imanishi and co-workers have observed CO gas from the carbon electrode in the LiCl and LiOH saturated aqueous solution above 3.3 V vs. Li+/Li [37]. The amount of CO evolved was significantly decreased by the addition of catalyst to the carbon electrode. The hydrophobic carbon produced less CO than the hydrophilic carbon as shown by Bruce etal. More recently, Bruce and co-workers [26] have demonstrated the cyclic performance of the non-aqueous lithium-oxygen cell with nanoporous gold electrodes without a carbon substrate and 0.1 M LiClO4 in DMSO and 95% capacity retention from cycles 1 to 100 at a current density of 500 mA g-gold-1. They claimed that although DMSO is not stable with bare lithium anodes, it could be used with protected lithium [26].

The behavior of the electrode reaction in non-aqueous lithium-oxygen cells is complex, and the stability of the electrolyte and the carbon is questionable, especially at a high current density. First, it is important to develop an acceptable non-aqueous electrolyte and oxygen electrode for the lithium-air batteries. Furthermore, conventional porous carbon air electrodes are unable to provide mAh g-1 and mAh cm-2 capacities and discharge rates at the magnitudes required for really high energy density batteries for EV applications. An effective air electrode would need to present a

FIGURE 5

Direct comparison of the CO2 evolution from decomposition of (a) Li CO3 (formed from decomposition of the electrolyte), (b) lithium carbonate, and (c) Li213CO3 (formed from decomposition of the carbon cathode when cycled in 0.5 M LiCO4 in DMSO and 0.5 M LiPF6 in tetraglyme. The number 2 and 5 correspond to data collected at the end of the second and fifth discharge, respectively. Figure from Ref. [36].

much shorter diffusion path for oxygen and offer the largest possible surface area for Li2O2 deposition. These requirements call for an open porous conductor structure using nanostructured materials [38].

Aqueous lithium-air system

Lithium reacts with water to produce LiOH and hydrogen gas; therefore, to avoid this parasitic corrosion reaction, most researchers on rechargeable lithium-air batteries have been focused on using an aprotic solvent as the electrolyte [1,7,39-41]. However, non-aqueous lithium-air batteries have some severe problems that still need to be addressed, such as lithium corrosion by water and CO2 when operated in air (short self-life) [27], decomposition of the electrolyte during the discharge and charge process (poor cyclic performance) [14,15,21] and high discharge and charge cell voltage difference (low energy conversion efficiency) [35,42] and high polarization of the air electrode during the charge process (low power density) [9,23,43]. Some of these problems could be removed with the use of an aqueous electrolyte and a water stable lithium electrode. Thus, the prerequisite to obtain a practical aqueous lithium-air system is to develop a water stable lithium electrode. The Li/aqueous electrolyte/O2 cell potential is 3.856 V vs. the standard oxygen electrode, and water decomposes spontaneously [2]. If lithium metal is in contact with a lithium conducting electrolyte that is stable at a potential lower than "3.040 V vs. NHE, then water is not decomposed. The protective layer should be stable in water. The concept of a water stable lithium electrode was proposed by Visco et al. in 2004 [44]. This electrode concept adopts a water stable NASICON-type lithium ion conducting solid electrolyte as a protective layer that covers and isolates the lithium metal from direct contact with the aqueous electrolyte. The NASI-CON-type lithium ion conducting solid electrolyte, Li1+xAxM2_x (PO4)3(A = Al, Sc, Y, M = Ti, Ge) was reported by Aono etal. [45,46]. Following this work, many researchers have studied this type of solid lithium conductor and the highest conductivity of 4.6 x 10"3Scm"1 was reported for glass ceramics of Li1+xAlx-Ge2"X(PO4)3 by Kumer et al. [47]. The lithium ion conductivity value is comparable to that of a non-aqueous lithium ion conducting electrolyte, because the lithium ion transport number of the solid electrolyte is unity, but that of the non-aqueous electrolyte is less than 0.5. The water stability of Li1+xAlxTi2_x(PO4)3 (LTAP) and Li1.4Al0.4Ge1.6(PO4)3 (LAGP) were examined by Ima-nishi et al. [48-50], where LTAP and LGAP were unstable in an aqueous solution of 1 M LiOH, but stable in an aqueous solution saturated with LiOH and LiCl. These results suggest that LATP and LAGP could be used as the protective layer for the water-stable lithium metal electrode in saturated LiOH/LiCl aqueous solution, because the reaction product of the aqueous lithium-air battery is LiOH. However, LATP and LAGP are unstable in direct contact with lithium metal. Therefore, an interface layer between lithium metal and LATP should be used. Visco et al. [44] have used a thin Li3N layer. Imanishi and co-workers proposed a polyethylene oxide (PEO) based polymer electrolyte [51]. Fig. 6 shows the lithium deposition and stripping potential change at 1.0 mA cm"2 and at 60 °C for Li/PEO 18LiTFSI-2TEGDME/LTAP/ saturated LiCl aqueous solution/Pt, air cell, where a platinum black reference electrode was used [52]. The composite polymer electrolyte of PEO18LiTFSI and TEGDME showed a low and stable interface resistance between lithium and the polymer electrolyte of 85 V cm2 at 60 °C. The lithium electrode is quite stable in the aqueous electrolyte. The solubility of LiOH is 5.3 ML"1 and saturation of LiOH in the electrolyte is reached at about 5% discharge depth. Stevens and co-workers proposed using an anion

FIGURE 6

Discharge and charge profiles for Li/PEO18LiTFSI-2TEGDME/LTAP/saturated

LiCl aqueous solution/Pt, air cell at 1.0 mAcm [52].

and 60 °C. Figure from Ref.

exchange membrane and an oxygen evolution electrode [53-55]. The schematic diagram of the proposed cell is shown in Fig. 7(a) and the cyclic performance at 0. 1 mA cm"2 is shown in Fig. 7(b). More than 100 cycles were obtained without any degradation when cycled with a low current density of 0.1 mA cm"2 for a short polarization period of 0.2 h, where a non-aqueous electrolyte was used as protective layer between the lithium anode and LTAP. The cell performance of the aqueous lithium-air cell with a limited amount of water (active materials) have been studied by Imanishi and co-workers, Fig. 8 shows the typical charge and discharge performance for the Li/LiClO4-EC-DME/LTAP/saturated LiOH and LiCl aqueous solution/carbon black, air cell at 0.88 mA cm"2 2 and at room temperature, where the platinum mesh with platinum black was used as the oxygen evolution electrode. The discharge capacity of 300 mAh g"1 (weight of water) corresponds to 30% utility of water. Flat discharge and charge potentials were obtained. The energy density was calculated to be

FIGURE 7

FIGURE 8

Potential profiles of the platinum black air electrode for Li/1 M LiClO4-EC-DME/LTAP/10 M LiCl-saturated LiOH/Pt, air at 0.88 mAcm"2 and 25 °C.

853Whkg"1 from the discharge capacity and OCV of 3.0 V. The capacity was more than two times higher than that of conventional lithium-ion batteries. The high discharge and charge over-potentials could be explained by the high cell resistance. The cell design should now be reconstructed with a limited amount of water.

Imanishi and co-workers reported successful charge and discharge performance using an aqueous electrolyte solution of acetic acid and lithium acetate as shown in Fig. 9 [56]. LTAP is stable in acetic acid (HAc) with a saturated lithium acetate (LiAc) aqueous solution [57]. The Li/PEO18LiTFSI-10 wt% BaTiO3/LTAP/HAc-H2O-LiAc/carbon, air cell shows an excellent cyclic performance at 60 °C and 0.5 mA cm"2. The discharge capacity was calculated as 250 mAh g-HAc"1 with consumption of 56% HAc, corresponding to an energy density of 779 Wh kg"1 against the weight of both lithium and HAc. This system can retain a discharge-charge capacity of 250 mAh g"1 for 15 cycles.

— ■ — Cycle 1

— •— Cycle 2

Cycle 5 —T— Cycle 10 —♦— Cycle 15

2 4 6 8 10 12 14 16 Cycle number _l_i_I_i_I_i_[_

0 25 50 75 100 125 150 175 200 225 250 275 Capacity / mAh g-1

FIGURE 9

Schematic diagram of the proposed device by Stevens et al. and cycling performance of the oxygen evolution electrode. Figure from Ref. [53].

Charge-discharge performance of Li/PEO18LiTFSI-10 wt% BaTiO3/HAc-H2O-LiAc(saturated)/carbon, air cell at 0.5 mA cm"2, at 60 °C. The amount of HAc in the electrolyte was 1 mg. The cell was sealed in a high pressure vessel with 3 atm of air to suppress evaporation of the electrolyte. Figure from Ref. [56].

Summary

In the last ten years, many materials for lithium-air battery systems have been reported by many researchers; especially non-aqueous electrolytes, water-stable solid electrolytes, and catalysts for Li2O2 reduction. However, the technical basis for practical high power density and extended deep cycling has yet to be demonstrated. Lithium-air battery research and technology is still in its initial stage. Some researchers are not optimistic for the future of lithium-air batteries, especially with respect to the volumetric energy density and power density. The capacity of air electrode in non-aqueous lithium-air batteries is dependent on the specific area of the carbon electrode [58,59]. High capacities of the oxygen electrode of more than 5000 mAh g-c"1 have been reported, where the current density was as lowas0.01 mA cm"1 [27] orthe packing density as low as 0.125 gmL"1 [20]. An air electrode capacity of 2000 mAh g"1 could be expected with moderate current density and packing density [60]. Lithium metal electrodes with flammable electrolytes could not be used for EV battery applications, for safety reasons. Therefore, candidates such as Li4.4Si with capacity of 2000 mAh g"1 are more promising [61]. For aqueous lithium-air batteries, the air electrode capacity is not dependent on the specific area of the carbon, but on the amount of aqueous electrolyte, assuming the reaction product of LiOHH2O is deposited in the electrolyte. The expected energy density for non-aqueous and aqueous lithium-air batteries could be acceptable for EV applications; however, such batteries must be capable of being discharged and charged at the rate required of EV applications, which may be the most challenging target yet.

References

[1] P.G. Bruce, et al. Nat. Mater. 11 (2012) 19.

[2] T. Zhang, et al. Chem. Lett. 40 (2011) 669.

[3] M.W. Chase Jr., J. Phys. Chem. Ref. Data, Monograph 9 (1990) 1510.

[4] T. Zhang, et al. Electrochem. Solid State Lett. 14 (2011) A45.

[5] N. Imanishi, O. Yamamoto, in: B. Scrosati, et al. (Eds.), Lithium Batteries: Advanced Technologies and Applications, John Wiley & Sons, Hoboken, NJ, USA, 2013, pp. 217-232.

[6] www.caranddriver.com/news/2013-nissan-leaf-photos-and-info-news.

[7] J. Christensen, et al. J. Electrochem. Soc. 159 (2012) R1.

[8] K.M. Abraham, Z. Jiang, J. Electrochem. Soc. 143 (1996) 1.

[9] T. Ogasawara, et al. J. Am. Chem. Soc. 128 (2006) 1390.

[10] B. Scrosati, in: M. Wakihara, O. Yammoto (Eds.), Lithium Ion Batteries, Fundamental and Performance, Kodansha, Tokyo, Japan, 1998, pp. 18-244.

[11] J. Read, J. Electrochem. Soc. 149 (2002) A1190.

[12 [13 [14 [15 [16 [17 [18 [19 [20] [21 [22] [23 [24 [25 [26 [27 [28 [29 [30] [31 [32] [33 [34 [35 [36 [37 [38 [39 [40] [41 [42] [43 [44

[45 [46 [47 [48 [49 [50 [51 [52 [53 [54 [55 [56 [57 [58 [59 [60] [61

J. Read, et al. J. Electrochem. Soc. 150 (2003) A1351. J. Read, J. Electrochem. Soc. 153 (2006) A96.

F. Mizuno, et al. Electrochemistry 78 (2010) 403.

S.A. Freunberger, et al. J. Am. Chem. Soc. 133 (2011) 8040. B.D. McCloskey, et al. J. Phys. Chem. C 116 (2012) 23897. B.D. McCloskey, et al. J. Phys. Chem. Lett. 3 (2012) 3034.

B.D. McCloskey, et al. J. Phys. Chem. Lett. 2 (2011) 1161.

G. Girishkumar, et al. J. Phys. Chem. Lett. 1 (2010) 2193.

C.O. Laoire, et al. J. Electrochem. Soc. 158 (2011) A302. S.A. Freubberg, et al. Angew. Chem. Int. Ed. 50 (2011) 1. J. Hassoun, et al. Angew. Chem. Int. Ed. 50 (2011) 2999.

H.-G. Jung, et al. Nat. Chem. 4 (2012) 579.

D. Aurbach, E. Granot, Electrochem. Acta 42 (1997) 697.

C.O. Laoire, et al. J. Phys. Chem. C 114 (2010) 9178. Z. Peng, et al. Science 337 (2012) 563.

T. Kuboki, et al. J. Power Sources 146 (2005) 766.

D. Zhang, et al. J. Power Sources 195 (2010) 1202. H. Nakamoto, et al. J. Power Sources 263 (2013) 19. B. Kumar, et al. J. Electrochem. Soc. 157 (2010) A50.

B. Kumar, J. Kumar, J. Electrochem. Soc. 157 (2010) A611.

J.S. Thokehom, et al. J. Electrohem. Soc. 155 (2008) A915.

V. Glordani, et al. Electrochem. Solid State Lett. 13 (2010) A180.

Y.-C. Lu, et al. Electrochem. Solid State Lett. 13 (2010) A60.

Y.-C. Lu, et al. J. Am. Chem. Soc. 132 (2010) 12170.

M.M. Ottakam, et al. J. Am. Chem. Soc. 135 (2013) 494.

H. Ohkuma, et al. J. Power Sources 245 (2014) 947.

J. Wang, Y. Li, X. Sun, Nano Energy 2 (2013) 433.

M. Armand, J.-M. Trascon, Nature 451 (2008) 652.

A. Kraytsberg, Y. Ein-Eli, J. Power Sources 196 (2011) 886.

D. Capson, et al. J. Power Sources 220 (2012) 253.

H. Cheng, K. Scott, Appl. Catal. B 140 (2011) 108-109.

U. Sahapatsombur, et al. J. Power Sources 227 (2013) 243.

S.J. Visco, et al. 12th International Meeting on Lithium Batteries, Abstract # 53,

Nara, Japan, 2004.

H. Aono, et al. J. Electrochem. Soc. 136 (1989) 590. H. Aono, et al. J. Electrochem. Soc. 137 (1990) 1023. J.S. Thockchom, et al. J. Power Sources 195 (2010) 2870. S. Hasegawa, et al. J. Power Sources 189 (2009) 371. Y. Shimonishi, et al. J. Power Sources 196 (2011) 5128. M. Zhang, et al. J. Electrochem. Soc. 159 (2012) A1114. T. Zhang, et al. J. Electrochem. Soc. 155 (2008) A965. H. Wang, et al. Membranes 3 (2013) 298. P. Stevens, et al. Electrochem. Soc. Trans 28 (2010) 1. L. Puech, et al. J. Power Sources 214 (2012) 330. F. Moureaux, et al. J. Power Sources 229 (2013) 123. T. Zhang, et al. Chem. Commun. 46 (2010) 1661. Y. Shimonishi, et al. J. Power Sources 195 (2010) 6187. P. Andrei, et al. J. Electrochem. Soc. 157 (2010) A1286. J.F. Zhang, et al. J. Electrochem. Soc. 155 (2008) A432. S.D. Beattie, et al. J. Electrochem. Soc. 156 (2009) A44. J. Cho, Mater. Chem. 26 (2010) 4009.