A novel high energy density rechargeable lithium/air battery

Abstract

A novel rechargeable lithium/air battery was fabricated, which consisted of a water-stable multilayer Li-metal anode, acetic acid–water electrolyte, and a fuel-cell analogous air-diffusion cathode and possessed a high energy density of 779 W h kg⁻¹, twice that of the conventional graphite/LiCoO₂ cell.

---

Lithium/air batteries are attractive energy storage systems in view of their high energy density for application in electric vehicles, the calculated energy density of which is comparable with that of gasoline. The first challenge for a rechargeable lithium/air battery is to ensure a reversible reaction. Such a battery, consisting of a thin Li metal foil anode and a carbon electrode with cobalt phthalocyanine as the catalyst, was initially reported to be recharged with good coulombic efficiency in 1996.¹ More recently, Bruce et al.² have presented high performance cells using a non-aqueous liquid electrolyte and a porous carbon electrode with electrolytic manganese dioxide as the catalyst. Although these systems are of considerable interest, their active life is limited by the diffusion of oxygen, carbon dioxide and water through the electrolyte and their subsequent reaction with lithium. To date, positive results have been obtained for these systems under pure oxygen or dry oxygen and nitrogen mixture.^1–3 However, for practical application in electrical vehicles, the air used will contain moisture. The electrochemical performance of lithium/air batteries is drastically reduced when the cell is cycled in air, even when a hydrophobic membrane is used to prevent the infiltration of moisture from the air. Thus, protection of the Li metal anode from water is the most critical point for long term stability of the lithium/air battery as pointed out by Armand and Tarascon.⁴ Furthermore, previously reported lithium/air batteries have employed organic electrolytes and the discharge reaction product of Li₂O₂ or Li₂O deposited on a carbon electrode. High charge–discharge polarization were observed in these cells and resulted in energy loss. Therefore, energy losses from charging and discharging should be reduced for applications such as electric vehicles.

Visco et al.⁵ have proposed a water-stable lithium electrode protected by NASICON-type water-stable solid electrolyte. However, we have observed that these solid electrolytes are unstable in strong basic and acidic media. The discharge/charge reaction of a lithium/air battery with an aqueous electrolyte solution is as follows:

2Li + 1/2O₂ + H₂O ↔ 2LiOH (1)

The concentration of OH⁻ in the electrolyte solution increases with increase in the discharge depth, which hinders the long term cycling of the water-stable lithium anode in aqueous media. Here we report the stable charge–discharge performance of a cell employing a solution of CH₃COOH (HOAc)–H₂O–CH₃COOLi (LiOAc) as electrolyte under ambient atmosphere. The NASICON-type Li_1.35Al_0.25Ti_1.75S_0.3P_2.7O₁₂ (LTAP) glass ceramic anode material used in this study is stable in the HOAc–H₂O–LiOAc solution.

The LTAP was supplied by Ohara Inc., Japan. The thickness of the LTAP plate was ca. 150 μm. The electrical conductivity of the LTAP plate at 60 °C was ca. 4 × 10⁻⁴ S cm⁻¹. The X-ray diffraction (XRD) features of the LATP plate are mainly due to the NASICON-type LiTi₂(PO₄)₃ phase (Fig. 1A). The LTAP plate was stable in neutral water with lithium ions, but was unstable in 1 M aqueous LiOH (Fig. 1B) and produced a high resistance phase, evidenced by a peak at 23° 2θ, which was indexed to be the Li₃PO₄ phase. LTAP powder was completely decomposed in 5 M aqueous HCl solution after 1 week and showed a very different diffraction pattern (Fig. 1D). Thus, the stability of LTAP in a weak acid of acetic acid–water solution was examined. The XRD pattern of LTAP powder immersed in the LiOAc saturated 90 vol% HOAc aqueous solution for 3 weeks at 50 °C (Fig. 1C) showed no change from that of the pristine LTAP, which indicated structural stability. The electrical conductivity of sintered plate samples with Au sputtered electrodes was measured using the AC impedance technique. LTAP immersed in pure HOAc for 3 weeks showed a significant decrease in conductivity; however, this decrease was suppressed by the addition of LiOAc salt in the HOAc solution. No change in conductivity was observed for LTAP immersed in the LiOAc saturated 90 vol% HOAc aqueous solution (pH = 3.34) at 50 °C for 5 months. The amount of Li-ions in solution and the electrical conductivity in the LiOAc saturated HOAc–H₂O solution increased with increasing H₂O content. The content of Li-ions in the LiOAc-saturated 90 vol% HOAc aqueous solution was 228 g L⁻¹ and the conductivity of the solution at room temperature was 2 × 10⁻³ S cm⁻¹. The conductivity is slightly lower than that of an electrolyte for lithium-ion batteries, but is acceptable for practical applications.

Fig. 1 XRD patterns of (A) pristine LTAP, and (B)–(D) LTAP immersed in various aqueous solutions at 50 °C: (B) in aqueous 1 M LiOH for 3 weeks, (C) in 90 vol% HOAc–10 vol% H₂O-saturated LiOAc for 3 weeks, (D) in aqueous 5 M HCl for one week. Powder XRD data was collected with a Rigaku RINT-2500HL diffractometer using Cu-Kα radiation and a step width of 0.03° 2θ.

The lithium/air battery proposed in this study is constructed of a lithium-metal active layer, a lithium-ion conducting polymer buffer-layer, a water-stable LTAP protective layer, the HOAc–H₂O–LiOAc electrolyte, and a carbon air cathode including superfine Pt particles (1–4 nm) as catalyst (Fig. 2) or a Pt-black only cathode (for Fig. 3). The Li/polymer electrolyte/LTAP anode was sealed using a plastic film, leaving a square window of 5 × 5 mm². The polymer electrolyte was used to prevent reaction between the lithium anode and LTAP, because the water-stable LTAP is unstable in direct contact with lithium metal and produces a highly resistive layer at the interface.⁶ The polymer electrolyte used was polyethylene oxide (PEO) with Li(CF₃SO₂)₂N (LiTFSI) (O/Li = 18/1). The method for preparation of the polymer electrolyte has been previously reported.⁶ The conductivity of the polymer electrolyte at 60 °C was ca. 3 × 10⁻⁴ S cm⁻¹.

Fig. 2 Schematic diagram of the proposed lithium–air system.

Fig. 3 Time dependence of Li electrode and oxygen electrode potential change of LiPEO₁₈LiTFSI|LTAP|HOAc (90 vol%)–H₂O–LiOAc (sat) (10 vol%)|Pt-black air cell at various constant current densities at 60 °C.

The cell resistance of the Li|PEO₁₈LiTFSI|LTAP|HOAc–H₂O–LiOAc (sat)|Pt-black air cell was 5466 Ω cm² at 25 °C and drastically decreased to 185 Ω cm² at 60 °C (see Fig. S1 in ESI†). The cell resistance is comparable to that of a cell with aqueous 1 M LiCl solution.⁷ According to the impedance analysis, the cell resistance is mainly due to the interface resistance between the lithium anode and polymer electrolyte and also between the polymer electrolyte and LTAP. The resistance between LTAP and the electrolyte solution was low. The open circuit voltage (OCV) of this cell was 3.69 V at room temperature, and this value remained constant for one month. The OCV is almost the same as that of the Li|PEO|LTAP|aqueous 1 M LiCl|Pt-black air cell (3.70 V).⁶ The following reversible cell reactions for the anode, cathode and total cell can be respectively expressed as:

2Li ↔ 2Li⁺ + 2e, (2)

1/2O₂ + 2CH₃COOH + 2e ↔ H₂O + 2CH₃COO⁻, (3)

2Li + 1/2O₂ + 2CH₃COOH ↔ 2CH₃COOLi + H₂O. (4)

The calculated OCV for the total reaction is 4.07 V at 25 °C, assuming that the activity of Li⁺ in the PEO electrolyte is unity and the oxygen partial pressure at the Pt electrode is 0.2 atm. The difference between the observed and experimental OCVs may be due to the low activity of Li-ion in PEO, the low oxygen partial pressure on the air electrode, and the junction potentials between the polymer electrolyte and LTAP, and also between LTAP and the HOAc–H₂O–LiOAc solution, because the polymer electrolyte, electrolyte solution, and LTAP have different conductivities and lithium ion transport numbers. The energy density of the Li/HOAc/air system was calculated to be 1.478 W h kg⁻¹ from the weight of lithium metal and HOAc, and the OCV of 3.69 V, which is approximately 3.6 times higher than that for the graphite/LiCoO₂ system (410 W h kg⁻¹).

The time dependence of the Li electrode potential and the oxygen electrode potential of the Li/PEO₁₈LiTFSI|LTAP|HOAc–H₂O–LiOAc (sat)|Pt-black air cell measured at various constant current drains at 60 °C and under open air are shown in the Fig. 3. The polarization at the lithium electrode and oxygen electrode is very low at a current density as high as 0.5 mA cm⁻² during the charge and discharge processes. Low electrode polarization is important to obtain high energy conversion efficiency for the rechargeable battery. Lithium/air cells with the organic electrolyte have high polarization during the charge/discharge process.² The high polarization may be due to the high activation energy required to reduce the Li₂O₂ reaction product during charge and to produce Li₂O₂ during discharge. Recently, an excellent air electrode catalyst was proposed by Bruce et al.,² although the polarization was still high. The charge process in our system is the oxidation of H₂O, as shown in eqn (3), which has been extensively studied in water electrolysis to be not so high at a current density of several mA cm⁻². The oxygen overpotential in 28% KOH solution at 70 °C was as low as 260 mV at 75 mA cm⁻² on NiCo₂O₄.⁸ The oxygen reduction reaction on a carbon electrode with a Pt-black catalyst in aqueous solution has been well known for fuel cells; the overpotential in 85% H₃PO₄ solution at 120 °C was only 200 mV at 10 mA cm⁻².⁹

The results shown in Fig. 3 were obtained for the test cell with an excess amount of the cathode active materials of HOAc. According to the cell reaction of Li/HOAc/air, the cell capacity depends on the amount of HOAc. In order to estimate the energy density of the proposed Li/HOAc/air system, it is important to investigate what amount of the active HOAc materials can be used. Cells with a defined amount of the LiOAc saturated HOAc (90 vol%) aqueous solution and an excess amount of lithium anode were constructed, where the HOAc solution was infiltrated in a porous polyethylene separator (100 μm thick), and a carbon sheet (320 μm thick) air electrode with a platinum catalyst was used as the air electrode (Fig. 2). The charge–discharge performance was examined at 0.5 mA cm⁻² and 60 °C under 3 atm of air (Fig. 4(a)). Flat discharge and charge potentials were obtained. A discharge capacity of 225 mA h g⁻¹ was obtained for 56% HOAc utilization. The estimated energy density from the weight of the lithium anode and HOAc is as high as 779 W h kg⁻¹. This energy density should be compared to that estimated from the electrode capacities and discharge cell voltage for a typical lithium-ion battery with a LiCoO₂ cathode and graphite anode (380 W h kg⁻¹),¹⁰ and also that estimated for a lithium-air battery with gel-polymer electrolytes (250–350 W h kg⁻¹) reported by Abraham et al.¹ The present result is two times higher than that of the graphite/LiCoO₂ cell. Therefore, the proposed lithium-air batteries are expected to have an energy density of more than 400 W h kg⁻¹ in practical applications, considering the practical energy density established for graphite/LiCoO₂ batteries is 200 W h kg⁻¹. The prototype lithium/air cell can retain a discharge–charge capacity of 250 mA h g⁻¹ (based on the weight of HOAc) within 15 cycles, as shown in Fig. 4(b). In this case, the HOAc utilization was the same as that in Fig. 4(a) and the cut-off voltage was set as 3.2 V. The potential difference between the discharge and charge platform in the second cycle was 0.75 V, which suggests low overpotential characteristics for the aqueous lithium/air battery.

Fig. 4 Charge–discharge performance of Li|PEO|LTAP|HOAc–H₂O–LiOAc (sat)|carbon air cell at 0.5 mA cm⁻² and 60 °C. (a) The volume ratio of the electrolyte was 90/10 HOAc/H₂O. (b) The amount of HOAc in the electrolyte was defined as 1 mg. The cell was sealed in a high pressure vessel with 3 atm of air to suppress evaporation of the electrolyte.

In summary, we have proposed a new type of rechargeable lithium-air battery with an expected energy density of more than 400 W h kg⁻¹. Optimization of the cell construction will be continued to improve the cycling performance. We are presently attempting to develop a battery with lower interface resistance between the lithium metal and PEO buffer-layer, an appropriate cathode structure with a good balance of hydrophobic and hydrophilic properties, and a new catalyst for the air electrode to obtain lower polarization.

This work was supported by the New Energy and Industrial Technology Development Organization (NEDO) of Japan under the project “Development of High Performance Battery System for Next-Generation Vehicles”. We thank Ohara Inc. for supplying the LTAP samples.

Notes and references

1. (a) K. M. Abraham and Z. Jiang, J. Electrochem. Soc., 1996, 143, 1; (b) M. R. Palacín, Chem. Soc. Rev., 2009, 38, 2565.
2. (a) T. Ogasawara, A. Débart, M. Holzapfel, P. Novak and P. G. Bruce, J. Am. Chem. Soc., 2006, 128, 1390; (b) A. Débart, A. J. Paterson, J. Bao and P. G. Bruce, Angew. Chem., Int. Ed., 2008, 47, 4521.
3. (a) J. Read, K. Mutolo, M. Ervin, W. Behl, J. Wolfenstine, A. Driedger and D. Foster, J. Electrochem. Soc., 2003, 150, A1351; (b) S. S. Sandhu, J. P. Fellner and G. W. Brutchen, J. Power Sources, 2007, 164, 365.
4. M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652.
5. S. J. Visco, E. Nimon, B. Katz, L. C. D. Jonghe and M. Y. Chu, Abstract 0389, The 210th Electrochemical Society Meeting Abstracts, Cancun, Mexico, 2006, Vol. 2006-2.
6. T. Zhang, N. Imanishi, S. Hasegawa, A. Hirano, J. Xie, Y. Takeda, O. Yamamoto and N. Sammes, J. Electrochem. Soc., 2008, 155, A965.
7. T. Zhang, N. Imanishi, S. Hasegawa, A. Hirano, J. Xie, Y. Takeda, O. Yamamoto and N. Sammes, Electrochem. Solid-State Lett., 2009, 12, A132.
8. R. L. LeRoy, M. B. Jonjua, R. Renaud and U. Leuenberger, J. Electrochem. Soc., 1979, 126, 1674.
9. W. Vogel, J. Lundquist and A. Bradford, Electrochim. Acta, 1972, 17, 1735.
10. (a) K. Mizushima, P. C. Jones, P. J. Wiseman and J. B. Goodenough, Solid State Ionics, 1981, 3–4, 171; (b) Z. Jiang, M. Alamgir and K. M. Abraham, J. Electrochem. Soc., 1995, 142, 333.

---

Footnote

† Electronic supplementary information (ESI) available: Fig. S1: Arrhenius plots and impedance spectra of the cell used for Fig. 3. See DOI: 10.1039/b920012f

---